text stringlengths 17 3.36M | source stringlengths 3 333 | __index_level_0__ int64 0 518k |
|---|---|---|
Solving constraint satisfaction problems (CSPs) is a notoriously expensive computational task. Recently, it has been proposed that efficient stochastic solvers can be obtained through appropriately configured spiking neural networks performing Markov Chain Monte Carlo (MCMC) sampling. The possibility to run such models on massively parallel, low-power neuromorphic hardware holds great promise; however, previously proposed networks are based on probabilistically spiking neurons, and thus rely on random number generators or external noise sources to achieve the necessary stochasticity, leading to significant overhead in the implementation. Here we show how stochasticity can be achieved by implementing deterministic models of integrate and fire neurons using subthreshold analog circuits that are affected by thermal noise. We present an efficient implementation of spike-based CSP solvers using a reconfigurable neural network VLSI device, and the device's intrinsic noise as a source of randomness. To illustrate the overall concept, we implement a generic Sudoku solver based on our approach and demonstrate its operation. We establish a link between the neuron parameters and the system dynamics, allowing for a simple temperature control mechanism. | Spiking Analog VLSI Neuron Assemblies as Constraint Satisfaction Problem
Solvers | 6,400 |
We improve the results by Siegelmann & Sontag (1995) by providing a novel and parsimonious constructive mapping between Turing Machines and Recurrent Artificial Neural Networks, based on recent developments of Nonlinear Dynamical Automata. The architecture of the resulting R-ANNs is simple and elegant, stemming from its transparent relation with the underlying NDAs. These characteristics yield promise for developments in machine learning methods and symbolic computation with continuous time dynamical systems. A framework is provided to directly program the R-ANNs from Turing Machine descriptions, in absence of network training. At the same time, the network can potentially be trained to perform algorithmic tasks, with exciting possibilities in the integration of approaches akin to Google DeepMind's Neural Turing Machines. | Turing Computation with Recurrent Artificial Neural Networks | 6,401 |
An important question in evolutionary computation is how good solutions evolutionary algorithms can produce. This paper aims to provide an analytic analysis of solution quality in terms of the relative approximation error, which is defined by the error between 1 and the approximation ratio of the solution found by an evolutionary algorithm. Since evolutionary algorithms are iterative methods, the relative approximation error is a function of generations. With the help of matrix analysis, it is possible to obtain an exact expression of such a function. In this paper, an analytic expression for calculating the relative approximation error is presented for a class of evolutionary algorithms, that is, (1+1) strictly elitist evolution algorithms. Furthermore, analytic expressions of the fitness value and the average convergence rate in each generation are also derived for this class of evolutionary algorithms. The approach is promising, and it can be extended to non-elitist or population-based algorithms too. | An Analytic Expression of Relative Approximation Error for a Class of
Evolutionary Algorithms | 6,402 |
Recent studies have shown that synaptic unreliability is a robust and sufficient mechanism for inducing the stochasticity observed in cortex. Here, we introduce Synaptic Sampling Machines, a class of neural network models that uses synaptic stochasticity as a means to Monte Carlo sampling and unsupervised learning. Similar to the original formulation of Boltzmann machines, these models can be viewed as a stochastic counterpart of Hopfield networks, but where stochasticity is induced by a random mask over the connections. Synaptic stochasticity plays the dual role of an efficient mechanism for sampling, and a regularizer during learning akin to DropConnect. A local synaptic plasticity rule implementing an event-driven form of contrastive divergence enables the learning of generative models in an on-line fashion. Synaptic sampling machines perform equally well using discrete-timed artificial units (as in Hopfield networks) or continuous-timed leaky integrate & fire neurons. The learned representations are remarkably sparse and robust to reductions in bit precision and synapse pruning: removal of more than 75% of the weakest connections followed by cursory re-learning causes a negligible performance loss on benchmark classification tasks. The spiking neuron-based synaptic sampling machines outperform existing spike-based unsupervised learners, while potentially offering substantial advantages in terms of power and complexity, and are thus promising models for on-line learning in brain-inspired hardware. | Stochastic Synapses Enable Efficient Brain-Inspired Learning Machines | 6,403 |
Recurrent Neural Networks (RNNs) have the ability to retain memory and learn data sequences. Due to the recurrent nature of RNNs, it is sometimes hard to parallelize all its computations on conventional hardware. CPUs do not currently offer large parallelism, while GPUs offer limited parallelism due to sequential components of RNN models. In this paper we present a hardware implementation of Long-Short Term Memory (LSTM) recurrent network on the programmable logic Zynq 7020 FPGA from Xilinx. We implemented a RNN with $2$ layers and $128$ hidden units in hardware and it has been tested using a character level language model. The implementation is more than $21\times$ faster than the ARM CPU embedded on the Zynq 7020 FPGA. This work can potentially evolve to a RNN co-processor for future mobile devices. | Recurrent Neural Networks Hardware Implementation on FPGA | 6,404 |
Evolutionary algorithms based on modeling the statistical dependencies (interactions) between the variables have been proposed to solve a wide range of complex problems. These algorithms learn and sample probabilistic graphical models able to encode and exploit the regularities of the problem. This paper investigates the effect of using probabilistic modeling techniques as a way to enhance the behavior of MOEA/D framework. MOEA/D is a decomposition based evolutionary algorithm that decomposes a multi-objective optimization problem (MOP) in a number of scalar single-objective subproblems and optimizes them in a collaborative manner. MOEA/D framework has been widely used to solve several MOPs. The proposed algorithm, MOEA/D using probabilistic Graphical Models (MOEA/D-GM) is able to instantiate both univariate and multi-variate probabilistic models for each subproblem. To validate the introduced framework algorithm, an experimental study is conducted on a multi-objective version of the deceptive function Trap5. The results show that the variant of the framework (MOEA/D-Tree), where tree models are learned from the matrices of the mutual information between the variables, is able to capture the structure of the problem. MOEA/D-Tree is able to achieve significantly better results than both MOEA/D using genetic operators and MOEA/D using univariate probability models, in terms of the approximation to the true Pareto front. | MOEA/D-GM: Using probabilistic graphical models in MOEA/D for solving
combinatorial optimization problems | 6,405 |
Particle swarm optimisation is a metaheuristic algorithm which finds reasonable solutions in a wide range of applied problems if suitable parameters are used. We study the properties of the algorithm in the framework of random dynamical systems which, due to the quasi-linear swarm dynamics, yields analytical results for the stability properties of the particles. Such considerations predict a relationship between the parameters of the algorithm that marks the edge between convergent and divergent behaviours. Comparison with simulations indicates that the algorithm performs best near this margin of instability. | Critical Parameters in Particle Swarm Optimisation | 6,406 |
This manuscript contains an outline of lectures course "Evolutionary Algorithms" read by the author. The course covers Canonic Genetic Algorithm and various other genetic algorithms as well as evolutionary strategies, genetic programming, tabu search and the class of evolutionary algorithms in general. Some facts, such as the Rotation Property of crossover, the Schemata Theorem, GA performance as a local search and "almost surely" convergence of evolutionary algorithms are given with complete proofs. The text is in Russian. | Evolutionary algorithms | 6,407 |
The rnn package provides components for implementing a wide range of Recurrent Neural Networks. It is built withing the framework of the Torch distribution for use with the nn package. The components have evolved from 3 iterations, each adding to the flexibility and capability of the package. All component modules inherit either the AbstractRecurrent or AbstractSequencer classes. Strong unit testing, continued backwards compatibility and access to supporting material are the principles followed during its development. The package is compared against existing implementations of two published papers. | rnn : Recurrent Library for Torch | 6,408 |
Today artificial neural networks are applied in various fields - engineering, data analysis, robotics. While they represent a successful tool for a variety of relevant applications, mathematically speaking they are still far from being conclusive. In particular, they suffer from being unable to find the best configuration possible during the training process (local minimum problem). In this paper, we focus on this issue and suggest a simple, but effective, post-learning strategy to allow the search for improved set of weights at a relatively small extra computational cost. Therefore, we introduce a novel technique based on analogy with quantum effects occurring in nature as a way to improve (and sometimes overcome) this problem. Several numerical experiments are presented to validate the approach. | On randomization of neural networks as a form of post-learning strategy | 6,409 |
This paper proposes an optimization algorithm based on how human fight and learn from each duelist. Since this algorithm is based on population, the proposed algorithm starts with an initial set of duelists. The duel is to determine the winner and loser. The loser learns from the winner, while the winner try their new skill or technique that may improve their fighting capabilities. A few duelists with highest fighting capabilities are called as champion. The champion train a new duelists such as their capabilities. The new duelist will join the tournament as a representative of each champion. All duelist are re-evaluated, and the duelists with worst fighting capabilities is eliminated to maintain the amount of duelists. Two optimization problem is applied for the proposed algorithm, together with genetic algorithm, particle swarm optimization and imperialist competitive algorithm. The results show that the proposed algorithm is able to find the better global optimum and faster iteration. | Duelist Algorithm: An Algorithm Inspired by How Duelist Improve Their
Capabilities in a Duel | 6,410 |
Oil refinery is one of industries that require huge energy consumption. The today technology advance requires energy saving. Heat integration is a method used to minimize the energy comsumption though the implementation of Heat Exchanger Network (HEN). CPT is one of types of Heat Exchanger Network (HEN) that functions to recover the heat in the flow of product or waste. HEN comprises a number of heat exchangers (HEs) that are serially connected. However, the presence of fouling in the heat exchanger has caused the decline of the performance of both heat exchangers and all heat exchanger networks. Fouling can not be avoided. However, it can be mitigated. In industry, periodic heat exchanger cleaning is the most effective and widely used mitigation technique. On the other side, a very frequent cleaning of heat exchanger can be much costly in maintenance and lost of production. In this way, an accurate optimization technique of cleaning schedule interval of heat exchanger is very essential. Commonly, this technique involves three elements: model to simulate the heat exchanger network, representative fouling model to describe the fouling behavior and suitable optimization algorithm to solve the problem of clening schedule interval for heat exchanger network. This paper describe the optimization of interval cleaning schedule of HEN within the 44-month period using PSO (particle swarm optimization). The number of iteration used to achieve the convergent is 100 iterations and the fitness value in PSO correlated with the amount of heat recovery, cleaning cost, and additional pumping cost. The saving after the optimization of cleaning schedule of HEN in this research achieved at $ 1.236 millions or 23% of maximum potential savings. | Cleaning Schedule Optimization of Heat Exchanger Networks Using Particle
Swarm Optimization | 6,411 |
Synapse plays an important role of learning in a neural network; the learning rules which modify the synaptic strength based on the timing difference between the pre- and post-synaptic spike occurrence is termed as Spike Time Dependent Plasticity (STDP). The most commonly used rule posits weight change based on time difference between one pre- and one post spike and is hence termed doublet STDP (DSTDP). However, D-STDP could not reproduce results of many biological experiments; a triplet STDP (T-STDP) that considers triplets of spikes as the fundamental unit has been proposed recently to explain these observations. This paper describes the compact implementation of a synapse using single floating-gate (FG) transistor that can store a weight in a nonvolatile manner and demonstrate the triplet STDP (T-STDP) learning rule by modifying drain voltages according to triplets of spikes. We describe a mathematical procedure to obtain control voltages for the FG device for T-STDP and also show measurement results from a FG synapse fabricated in TSMC 0.35um CMOS process to support the theory. Possible VLSI implementation of drain voltage waveform generator circuits are also presented with simulation results. | Triplet Spike Time Dependent Plasticity: A floating-gate Implementation | 6,412 |
In this article, we propose a novel Winner-Take-All (WTA) architecture employing neurons with nonlinear dendrites and an online unsupervised structural plasticity rule for training it. Further, to aid hardware implementations, our network employs only binary synapses. The proposed learning rule is inspired by spike time dependent plasticity (STDP) but differs for each dendrite based on its activation level. It trains the WTA network through formation and elimination of connections between inputs and synapses. To demonstrate the performance of the proposed network and learning rule, we employ it to solve two, four and six class classification of random Poisson spike time inputs. The results indicate that by proper tuning of the inhibitory time constant of the WTA, a trade-off between specificity and sensitivity of the network can be achieved. We use the inhibitory time constant to set the number of subpatterns per pattern we want to detect. We show that while the percentage of successful trials are 92%, 88% and 82% for two, four and six class classification when no pattern subdivisions are made, it increases to 100% when each pattern is subdivided into 5 or 10 subpatterns. However, the former scenario of no pattern subdivision is more jitter resilient than the later ones. | An Online Unsupervised Structural Plasticity Algorithm for Spiking
Neural Networks | 6,413 |
The paper is devoted to upper bounds on run-time of Non-Elitist Genetic Algorithms until some target subset of solutions is visited for the first time. In particular, we consider the sets of optimal solutions and the sets of local optima as the target subsets. Previously known upper bounds are improved by means of drift analysis. Finally, we propose conditions ensuring that a Non-Elitist Genetic Algorithm efficiently finds approximate solutions with constant approximation ratio on the class of combinatorial optimization problems with guaranteed local optima (GLO). | Level-Based Analysis of Genetic Algorithms for Combinatorial
Optimization | 6,414 |
The application of evolution in the digital realm, with the goal of creating artificial intelligence and artificial life, has a history as long as that of the digital computer itself. We illustrate the intertwined history of these ideas, starting with the early theoretical work of John von Neumann and the pioneering experimental work of Nils Aall Barricelli. We argue that evolutionary thinking and artificial life will continue to play an integral role in the future development of the digital world. | Digital Genesis: Computers, Evolution and Artificial Life | 6,415 |
NM-landscapes have been recently introduced as a class of tunable rugged models. They are a subset of the general interaction models where all the interactions are of order less or equal $M$. The Boltzmann distribution has been extensively applied in single-objective evolutionary algorithms to implement selection and study the theoretical properties of model-building algorithms. In this paper we propose the combination of the multi-objective NM-landscape model and the Boltzmann distribution to obtain Pareto-front approximations. We investigate the joint effect of the parameters of the NM-landscapes and the probabilistic factorizations in the shape of the Pareto front approximations. | Computing factorized approximations of Pareto-fronts using
mNM-landscapes and Boltzmann distributions | 6,416 |
We review several of the most widely used techniques for training recurrent neural networks to approximate dynamical systems, then describe a novel algorithm for this task. The algorithm is based on an earlier theoretical result that guarantees the quality of the network approximation. We show that a feedforward neural network can be trained on the vector field representation of a given dynamical system using backpropagation, then recast, using matrix manipulations, as a recurrent network that replicates the original system's dynamics. After detailing this algorithm and its relation to earlier approaches, we present numerical examples that demonstrate its capabilities. One of the distinguishing features of our approach is that both the original dynamical systems and the recurrent networks that simulate them operate in continuous time. | Synthesis of recurrent neural networks for dynamical system simulation | 6,417 |
The differential evolution (DE) algorithm suffers from high computational time due to slow nature of evaluation. In contrast, micro-DE (MDE) algorithms employ a very small population size, which can converge faster to a reasonable solution. However, these algorithms are vulnerable to a premature convergence as well as to high risk of stagnation. In this paper, MDE algorithm with vectorized random mutation factor (MDEVM) is proposed, which utilizes the small size population benefit while empowers the exploration ability of mutation factor through randomizing it in the decision variable level. The idea is supported by analyzing mutation factor using Monte-Carlo based simulations. To facilitate the usage of MDE algorithms with very-small population sizes, new mutation schemes for population sizes less than four are also proposed. Furthermore, comprehensive comparative simulations and analysis on performance of the MDE algorithms over various mutation schemes, population sizes, problem types (i.e. uni-modal, multi-modal, and composite), problem dimensionalities, and mutation factor ranges are conducted by considering population diversity analysis for stagnation and trapping in local optimum situations. The studies are conducted on 28 benchmark functions provided for the IEEE CEC-2013 competition. Experimental results demonstrate high performance and convergence speed of the proposed MDEVM algorithm. | Diversity Enhancement for Micro-Differential Evolution | 6,418 |
In this paper, we propose a novel neural network structure, namely \emph{feedforward sequential memory networks (FSMN)}, to model long-term dependency in time series without using recurrent feedback. The proposed FSMN is a standard fully-connected feedforward neural network equipped with some learnable memory blocks in its hidden layers. The memory blocks use a tapped-delay line structure to encode the long context information into a fixed-size representation as short-term memory mechanism. We have evaluated the proposed FSMNs in several standard benchmark tasks, including speech recognition and language modelling. Experimental results have shown FSMNs significantly outperform the conventional recurrent neural networks (RNN), including LSTMs, in modeling sequential signals like speech or language. Moreover, FSMNs can be learned much more reliably and faster than RNNs or LSTMs due to the inherent non-recurrent model structure. | Feedforward Sequential Memory Networks: A New Structure to Learn
Long-term Dependency | 6,419 |
Cortical Learning Algorithms based on the Hierarchical Temporal Memory, HTM have been developed by Numenta Incorporation from which variations and modifications are currently being investigated upon. HTM offers better promises as a future computational model of the neocortex the seat of intelligence in the brain. Currently, intelligent agents are embedded in almost every modern day electronic system found in homes, offices and industries worldwide. In this paper, we present a first step in realising useful HTM like applications specifically for mining a synthetic and real time dataset based on a novel intelligent agent framework, and demonstrate how a modified version of this very important computational technique will lead to improved recognition. | An Improved Intelligent Agent for Mining Real-Time Databases Using
Modified Cortical Learning Algorithms | 6,420 |
In case of decision making problems, classification of pattern is a complex and crucial task. Pattern classification using multilayer perceptron (MLP) trained with back propagation learning becomes much complex with increase in number of layers, number of nodes and number of epochs and ultimate increases computational time [31]. In this paper, an attempt has been made to use fuzzy MLP and its learning algorithm for pattern classification. The time and space complexities of the algorithm have been analyzed. A training performance comparison has been carried out between MLP and the proposed fuzzy-MLP model by considering six cases. Results are noted against different learning rates ranging from 0 to 1. A new performance evaluation factor 'convergence gain' has been introduced. It is observed that the number of epochs drastically reduced and performance increased compared to MLP. The average and minimum gain has been found to be 93% and 75% respectively. The best gain is found to be 95% and is obtained by setting the learning rate to 0.55. | A Fuzzy MLP Approach for Non-linear Pattern Classification | 6,421 |
In recent years the field of neuromorphic low-power systems that consume orders of magnitude less power gained significant momentum. However, their wider use is still hindered by the lack of algorithms that can harness the strengths of such architectures. While neuromorphic adaptations of representation learning algorithms are now emerging, efficient processing of temporal sequences or variable length-inputs remain difficult. Recurrent neural networks (RNN) are widely used in machine learning to solve a variety of sequence learning tasks. In this work we present a train-and-constrain methodology that enables the mapping of machine learned (Elman) RNNs on a substrate of spiking neurons, while being compatible with the capabilities of current and near-future neuromorphic systems. This "train-and-constrain" method consists of first training RNNs using backpropagation through time, then discretizing the weights and finally converting them to spiking RNNs by matching the responses of artificial neurons with those of the spiking neurons. We demonstrate our approach by mapping a natural language processing task (question classification), where we demonstrate the entire mapping process of the recurrent layer of the network on IBM's Neurosynaptic System "TrueNorth", a spike-based digital neuromorphic hardware architecture. TrueNorth imposes specific constraints on connectivity, neural and synaptic parameters. To satisfy these constraints, it was necessary to discretize the synaptic weights and neural activities to 16 levels, and to limit fan-in to 64 inputs. We find that short synaptic delays are sufficient to implement the dynamical (temporal) aspect of the RNN in the question classification task. The hardware-constrained model achieved 74% accuracy in question classification while using less than 0.025% of the cores on one TrueNorth chip, resulting in an estimated power consumption of ~17 uW. | Conversion of Artificial Recurrent Neural Networks to Spiking Neural
Networks for Low-power Neuromorphic Hardware | 6,422 |
We report about probabilistic likelihood estimates that are performed on time series using an echo state network with orthogonal recurrent connectivity. The results from tests using synthetic stochastic input time series with temporal inference indicate that the capability of the network to infer depends on the balance between input strength and recurrent activity. This balance has an influence on the network with regard to the quality of inference from the short term input history versus inference that accounts for influences that date back a long time. Sensitivity of such networks against noise and the finite accuracy of network states in the recurrent layer are investigated. In addition, a measure based on mutual information between the output time series and the reservoir is introduced. Finally, different types of recurrent connectivity are evaluated. Orthogonal matrices show the best results of all investigated connectivity types overall, but also in the way how the network performance scales with the size of the recurrent layer. | Orthogonal Echo State Networks and stochastic evaluations of likelihoods | 6,423 |
We present the soft exponential activation function for artificial neural networks that continuously interpolates between logarithmic, linear, and exponential functions. This activation function is simple, differentiable, and parameterized so that it can be trained as the rest of the network is trained. We hypothesize that soft exponential has the potential to improve neural network learning, as it can exactly calculate many natural operations that typical neural networks can only approximate, including addition, multiplication, inner product, distance, polynomials, and sinusoids. | A continuum among logarithmic, linear, and exponential functions, and
its potential to improve generalization in neural networks | 6,424 |
We present a spike-based unsupervised regenerative learning scheme to train Spiking Deep Networks (SpikeCNN) for object recognition problems using biologically realistic leaky integrate-and-fire neurons. The training methodology is based on the Auto-Encoder learning model wherein the hierarchical network is trained layer wise using the encoder-decoder principle. Regenerative learning uses spike-timing information and inherent latencies to update the weights and learn representative levels for each convolutional layer in an unsupervised manner. The features learnt from the final layer in the hierarchy are then fed to an output layer. The output layer is trained with supervision by showing a fraction of the labeled training dataset and performs the overall classification of the input. Our proposed methodology yields 0.92%/29.84% classification error on MNIST/CIFAR10 datasets which is comparable with state-of-the-art results. The proposed methodology also introduces sparsity in the hierarchical feature representations on account of event-based coding resulting in computationally efficient learning. | Unsupervised Regenerative Learning of Hierarchical Features in Spiking
Deep Networks for Object Recognition | 6,425 |
While classical neural networks take a position of a leading method in the machine learning community, spiking neuromorphic systems bring attention and large projects in neuroscience. Spiking neural networks were shown to be able to substitute networks of classical neurons in applied tasks. This work explores recent hardware designs focusing on perspective applications (like convolutional neural networks) for both neuron types from the energy efficiency side to analyse whether there is a possibility for spiking neuromorphic hardware to grow up for a wider use. Our comparison shows that spiking hardware is at least on the same level of energy efficiency or even higher than non-spiking on a level of basic operations. However, on a system level, spiking systems are outmatched and consume much more energy due to inefficient data representation with a long series of spikes. If spike-driven applications, minimizing an amount of spikes, are developed, spiking neural systems may reach the energy efficiency level of classical neural systems. However, in the near future, both type of neuromorphic systems may benefit from emerging memory technologies, minimizing the energy consumption of computation and memory for both neuron types. That would make infrastructure and data transfer energy dominant on the system level. We expect that spiking neurons have some benefits, which would allow achieving better energy results. Still the problem of an amount of spikes will still be the major bottleneck for spiking hardware systems. | Computing with hardware neurons: spiking or classical? Perspectives of
applied Spiking Neural Networks from the hardware side | 6,426 |
Training Brain Computer Interface (BCI) systems to understand the intention of a subject through Electroencephalogram (EEG) data currently requires multiple training sessions with a subject in order to develop the necessary expertise to distinguish signals for different tasks. Conventionally the task of training the subject is done by introducing a training and calibration stage during which some feedback is presented to the subject. This training session can take several hours which is not appropriate for on-line EEG-based BCI systems. An alternative approach is to use previous recording sessions of the same person or some other subjects that performed the same tasks (subject transfer) for training the classifiers. The main aim of this study is to generate a methodology that allows the use of data from other subjects while reducing the dimensions of the data. The study investigates several possibilities for reducing the necessary training and calibration period in subjects and the classifiers and addresses the impact of i) evolutionary subject transfer and ii) adapting previously trained methods (retraining) using other subjects data. Our results suggest reduction to 40% of target subject data is sufficient for training the classifier. Our results also indicate the superiority of the approaches that incorporated evolutionary subject transfer and highlights the feasibility of adapting a system trained on other subjects. | Reducing training requirements through evolutionary based dimension
reduction and subject transfer | 6,427 |
With this paper, we contribute to the growing research area of feature-based analysis of bio-inspired computing. In this research area, problem instances are classified according to different features of the underlying problem in terms of their difficulty of being solved by a particular algorithm. We investigate the impact of different sets of evolved instances for building prediction models in the area of algorithm selection. Building on the work of Poursoltan and Neumann [11,10], we consider how evolved instances can be used to predict the best performing algorithm for constrained continuous optimisation from a set of bio-inspired computing methods, namely high performing variants of differential evolution, particle swarm optimization, and evolution strategies. Our experimental results show that instances evolved with a multi-objective approach in combination with random instances of the underlying problem allow to build a model that accurately predicts the best performing algorithm for a wide range of problem instances. | A Feature-Based Prediction Model of Algorithm Selection for Constrained
Continuous Optimisation | 6,428 |
We investigate a new method to augment recurrent neural networks with extra memory without increasing the number of network parameters. The system has an associative memory based on complex-valued vectors and is closely related to Holographic Reduced Representations and Long Short-Term Memory networks. Holographic Reduced Representations have limited capacity: as they store more information, each retrieval becomes noisier due to interference. Our system in contrast creates redundant copies of stored information, which enables retrieval with reduced noise. Experiments demonstrate faster learning on multiple memorization tasks. | Associative Long Short-Term Memory | 6,429 |
The advantage of recurrent neural networks (RNNs) in learning dependencies between time-series data has distinguished RNNs from other deep learning models. Recently, many advances are proposed in this emerging field. However, there is a lack of comprehensive review on memory models in RNNs in the literature. This paper provides a fundamental review on RNNs and long short term memory (LSTM) model. Then, provides a surveys of recent advances in different memory enhancements and learning techniques for capturing long term dependencies in RNNs. | Learning Over Long Time Lags | 6,430 |
Timetabling is a problem faced in all higher education institutions. The International Timetabling Competition (ITC) has published a dataset that can be used to test the quality of methods used to solve this problem. A number of meta-heuristic approaches have obtained good results when tested on the ITC dataset, however few have used the ant colony optimization technique, particularly on the ITC 2007 curriculum based university course timetabling problem. This study describes an ant system that solves the curriculum based university course timetabling problem and the quality of the algorithm is tested on the ITC 2007 dataset. The ant system was able to find feasible solutions in all instances of the dataset and close to optimal solutions in some instances. The ant system performs better than some published approaches, however results obtained are not as good as those obtained by the best published approaches. This study may be used as a benchmark for ant based algorithms that solve the curriculum based university course timetabling problem. | Greedy Ants Colony Optimization Strategy for Solving the Curriculum
Based University Course Timetabling Problem | 6,431 |
Restricted Boltzmann Machines and Deep Belief Networks have been successfully used in probabilistic generative model applications such as image occlusion removal, pattern completion and motion synthesis. Generative inference in such algorithms can be performed very efficiently on hardware using a Markov Chain Monte Carlo procedure called Gibbs sampling, where stochastic samples are drawn from noisy integrate and fire neurons implemented on neuromorphic substrates. Currently, no satisfactory metrics exist for evaluating the generative performance of such algorithms implemented on high-dimensional data for neuromorphic platforms. This paper demonstrates the application of nonparametric goodness-of-fit testing to both quantify the generative performance as well as provide decision-directed criteria for choosing the parameters of the neuromorphic Gibbs sampler and optimizing usage of hardware resources used during sampling. | A Nonparametric Framework for Quantifying Generative Inference on
Neuromorphic Systems | 6,432 |
Firefly algorithm is a swarm based metaheuristic algorithm inspired by the flashing behavior of fireflies. It is an effective and an easy to implement algorithm. It has been tested on different problems from different disciplines and found to be effective. Even though the algorithm is proposed for optimization problems with continuous variables, it has been modified and used for problems with non-continuous variables, including binary and integer valued problems. In this paper a detailed review of this modifications of firefly algorithm for problems with non-continuous variables will be discussed. The strength and weakness of the modifications along with possible future works will be presented. | Firefly Algorithm for optimization problems with non-continuous
variables: A Review and Analysis | 6,433 |
We introduce an algorithm to do backpropagation on a spiking network. Our network is "spiking" in the sense that our neurons accumulate their activation into a potential over time, and only send out a signal (a "spike") when this potential crosses a threshold and the neuron is reset. Neurons only update their states when receiving signals from other neurons. Total computation of the network thus scales with the number of spikes caused by an input rather than network size. We show that the spiking Multi-Layer Perceptron behaves identically, during both prediction and training, to a conventional deep network of rectified-linear units, in the limiting case where we run the spiking network for a long time. We apply this architecture to a conventional classification problem (MNIST) and achieve performance very close to that of a conventional Multi-Layer Perceptron with the same architecture. Our network is a natural architecture for learning based on streaming event-based data, and is a stepping stone towards using spiking neural networks to learn efficiently on streaming data. | Deep Spiking Networks | 6,434 |
Multilayered artificial neural networks (ANN) have found widespread utility in classification and recognition applications. The scale and complexity of such networks together with the inadequacies of general purpose computing platforms have led to a significant interest in the development of efficient hardware implementations. In this work, we focus on designing energy efficient on-chip storage for the synaptic weights. In order to minimize the power consumption of typical digital CMOS implementations of such large-scale networks, the digital neurons could be operated reliably at scaled voltages by reducing the clock frequency. On the contrary, the on-chip synaptic storage designed using a conventional 6T SRAM is susceptible to bitcell failures at reduced voltages. However, the intrinsic error resiliency of NNs to small synaptic weight perturbations enables us to scale the operating voltage of the 6TSRAM. Our analysis on a widely used digit recognition dataset indicates that the voltage can be scaled by 200mV from the nominal operating voltage (950mV) for practically no loss (less than 0.5%) in accuracy (22nm predictive technology). Scaling beyond that causes substantial performance degradation owing to increased probability of failures in the MSBs of the synaptic weights. We, therefore propose a significance driven hybrid 8T-6T SRAM, wherein the sensitive MSBs are stored in 8T bitcells that are robust at scaled voltages due to decoupled read and write paths. In an effort to further minimize the area penalty, we present a synaptic-sensitivity driven hybrid memory architecture consisting of multiple 8T-6T SRAM banks. Our circuit to system-level simulation framework shows that the proposed synaptic-sensitivity driven architecture provides a 30.91% reduction in the memory access power with a 10.41% area overhead, for less than 1% loss in the classification accuracy. | Significance Driven Hybrid 8T-6T SRAM for Energy-Efficient Synaptic
Storage in Artificial Neural Networks | 6,435 |
Large-scale artificial neural networks have shown significant promise in addressing a wide range of classification and recognition applications. However, their large computational requirements stretch the capabilities of computing platforms. The fundamental components of these neural networks are the neurons and its synapses. The core of a digital hardware neuron consists of multiplier, accumulator and activation function. Multipliers consume most of the processing energy in the digital neurons, and thereby in the hardware implementations of artificial neural networks. We propose an approximate multiplier that utilizes the notion of computation sharing and exploits error resilience of neural network applications to achieve improved energy consumption. We also propose Multiplier-less Artificial Neuron (MAN) for even larger improvement in energy consumption and adapt the training process to ensure minimal degradation in accuracy. We evaluated the proposed design on 5 recognition applications. The results show, 35% and 60% reduction in energy consumption, for neuron sizes of 8 bits and 12 bits, respectively, with a maximum of ~2.83% loss in network accuracy, compared to a conventional neuron implementation. We also achieve 37% and 62% reduction in area for a neuron size of 8 bits and 12 bits, respectively, under iso-speed conditions. | Multiplier-less Artificial Neurons Exploiting Error Resiliency for
Energy-Efficient Neural Computing | 6,436 |
Convolutional neural networks (CNNs) are the cutting edge model for supervised machine learning in computer vision. In recent years CNNs have outperformed traditional approaches in many computer vision tasks such as object detection, image classification and face recognition. CNNs are vulnerable to overfitting, and a lot of research focuses on finding regularization methods to overcome it. One approach is designing task specific models based on prior knowledge. Several works have shown that properties of natural images can be easily captured using complex numbers. Motivated by these works, we present a variation of the CNN model with complex valued input and weights. We construct the complex model as a generalization of the real model. Lack of order over the complex field raises several difficulties both in the definition and in the training of the network. We address these issues and suggest possible solutions. The resulting model is shown to be a restricted form of a real valued CNN with twice the parameters. It is sensitive to phase structure, and we suggest it serves as a regularized model for problems where such structure is important. This suggestion is verified empirically by comparing the performance of a complex and a real network in the problem of cell detection. The two networks achieve comparable results, and although the complex model is hard to train, it is significantly less vulnerable to overfitting. We also demonstrate that the complex network detects meaningful phase structure in the data. | On Complex Valued Convolutional Neural Networks | 6,437 |
Quality assurance in production line demands reliable weld joints. Human made errors is a major cause of faulty production. Promptly Identifying errors in the weld while welding is in progress will decrease the post inspection cost spent on the welding process. Electrical parameters generated during welding, could able to characterize the process efficiently. Parameter values are collected using high speed data acquisition system. Time series analysis tasks such as filtering, pattern recognition etc. are performed over the collected data. Filtering removes the unwanted noisy signal components and pattern recognition task segregate error patterns in the time series based upon similarity, which is performed by Self Organized mapping clustering algorithm. Welder quality is thus compared by detecting and counting number of error patterns appeared in his parametric time series. Moreover, Self Organized mapping algorithm provides the database in which patterns are segregated into two classes either desirable or undesirable. Database thus generated is used to train the classification algorithms, and thereby automating the real time error detection task. Multi Layer Perceptron and Radial basis function are the two classification algorithms used, and their performance has been compared based on metrics such as specificity, sensitivity, accuracy and time required in training. | Real time error detection in metal arc welding process using Artificial
Neural Netwroks | 6,438 |
We present a digital implementation of the Spike Timing Dependent Plasticity (STDP) learning rule. The proposed digital implementation consists of an exponential decay generator array and a STDP adaptor array. On the arrival of a pre- and post-synaptic spike, the STDP adaptor will send a digital spike to the decay generator. The decay generator will then generate an exponential decay, which will be used by the STDP adaptor to perform the weight adaption. The exponential decay, which is computational expensive, is efficiently implemented by using a novel stochastic approach, which we analyse and characterise here. We use a time multiplexing approach to achieve 8192 (8k) virtual STDP adaptors and decay generators with only one physical implementation of each. We have validated our stochastic STDP approach with measurement results of a balanced excitation/inhibition experiment. Our stochastic approach is ideal for implementing the STDP learning rule in large-scale spiking neural networks running in real time. | A Stochastic Approach to STDP | 6,439 |
Online parameter controllers for evolutionary algorithms adjust values of parameters during the run of an evolutionary algorithm. Recently a new efficient parameter controller based on reinforcement learning was proposed by Karafotias et al. In this method ranges of parameters are discretized into several intervals before the run. However, performing adaptive discretization during the run may increase efficiency of an evolutionary algorithm. Aleti et al. proposed another efficient controller with adaptive discretization. In the present paper we propose a parameter controller based on reinforcement learning with adaptive discretization. The proposed controller is compared with the existing parameter adjusting methods on several test problems using different configurations of an evolutionary algorithm. For the test problems, we consider four continuous functions, namely the sphere function, the Rosenbrock function, the Levi function and the Rastrigin function. Results show that the new controller outperforms the other controllers on most of the considered test problems. | Adaptive Parameter Selection in Evolutionary Algorithms by Reinforcement
Learning with Dynamic Discretization of Parameter Range | 6,440 |
A draft memory model (DM) for neural networks with spike propagation delay (SNNwD) is described. Novelty in this approach are that the DM learns immediately, with stimuli presented once, without synaptic weight changes, and without external learning algorithm. Basal on this model is to trap spikes within neural loops. In order to construct the DM we developed two functional blocks, also described herein. The decoder block receives input from a single spikes source and connect it to one among many outputs. The selector block operates in the opposite direction, receiving many spikes sources and connecting one of them to a single output. We realized conceptual proofs by testing the DM in the prime numbers classifying task. This activation-based memory can be used as immediate and short-term memory. | A Draft Memory Model on Spiking Neural Assemblies | 6,441 |
Deep networks are now able to achieve human-level performance on a broad spectrum of recognition tasks. Independently, neuromorphic computing has now demonstrated unprecedented energy-efficiency through a new chip architecture based on spiking neurons, low precision synapses, and a scalable communication network. Here, we demonstrate that neuromorphic computing, despite its novel architectural primitives, can implement deep convolution networks that i) approach state-of-the-art classification accuracy across 8 standard datasets, encompassing vision and speech, ii) perform inference while preserving the hardware's underlying energy-efficiency and high throughput, running on the aforementioned datasets at between 1200 and 2600 frames per second and using between 25 and 275 mW (effectively > 6000 frames / sec / W) and iii) can be specified and trained using backpropagation with the same ease-of-use as contemporary deep learning. For the first time, the algorithmic power of deep learning can be merged with the efficiency of neuromorphic processors, bringing the promise of embedded, intelligent, brain-inspired computing one step closer. | Convolutional Networks for Fast, Energy-Efficient Neuromorphic Computing | 6,442 |
In this paper Hybrid Ant Colony Optimization (HAntCO) approach in solving Multi--Skill Resource Constrained Project Scheduling Problem (MS--RCPSP) has been presented. We have proposed hybrid approach that links classical heuristic priority rules for project scheduling with Ant Colony Optimization (ACO). Furthermore, a novel approach for updating pheromone value has been proposed, based on both the best and worst solutions stored by ants. The objective of this paper is to research the usability and robustness of ACO and its hybrids with priority rules in solving MS--RCPSP. Experiments have been performed using artificially created dataset instances, based on real--world ones. We published those instances that can be used as a benchmark. Presented results show that ACO--based hybrid method is an efficient approach. More directed search process by hybrids makes this approach more stable and provides mostly better results than classical ACO. | Hybrid Ant Colony Optimization in solving Multi-Skill
Resource-Constrained Project Scheduling Problem | 6,443 |
This paper introduces Adaptive Computation Time (ACT), an algorithm that allows recurrent neural networks to learn how many computational steps to take between receiving an input and emitting an output. ACT requires minimal changes to the network architecture, is deterministic and differentiable, and does not add any noise to the parameter gradients. Experimental results are provided for four synthetic problems: determining the parity of binary vectors, applying binary logic operations, adding integers, and sorting real numbers. Overall, performance is dramatically improved by the use of ACT, which successfully adapts the number of computational steps to the requirements of the problem. We also present character-level language modelling results on the Hutter prize Wikipedia dataset. In this case ACT does not yield large gains in performance; however it does provide intriguing insight into the structure of the data, with more computation allocated to harder-to-predict transitions, such as spaces between words and ends of sentences. This suggests that ACT or other adaptive computation methods could provide a generic method for inferring segment boundaries in sequence data. | Adaptive Computation Time for Recurrent Neural Networks | 6,444 |
Dataflow matrix machines are a powerful generalization of recurrent neural networks. They work with multiple types of arbitrary linear streams, multiple types of powerful neurons, and allow to incorporate higher-order constructions. We expect them to be useful in machine learning and probabilistic programming, and in the synthesis of dynamic systems and of deterministic and probabilistic programs. | Dataflow Matrix Machines as a Generalization of Recurrent Neural
Networks | 6,445 |
In this research, feedforward ANN (Artificial Neural Network) model is developed and validated for predicting the pH at 10 different locations of the distribution system of drinking water of Hyderabad city. The developed model is MLP (Multilayer Perceptron) with back propagation algorithm.The data for the training and testing of the model are collected through an experimental analysis on weekly basis in a routine examination for maintaining the quality of drinking water in the city. 17 parameters are taken into consideration including pH. These all parameters are taken as input variables for the model and then pH is predicted for 03 phases;raw water of river Indus,treated water in the treatment plants and then treated water in the distribution system of drinking water. The training and testing results of this model reveal that MLP neural networks are exceedingly extrapolative for predicting the pH of river water, untreated and treated water at all locations of the distribution system of drinking water of Hyderabad city. The optimum input and output weights are generated with minimum MSE (Mean Square Error) < 5%.Experimental, predicted and tested values of pH are plotted and the effectiveness of the model is determined by calculating the coefficient of correlation (R2=0.999) of trained and tested results. | pH Prediction by Artificial Neural Networks for the Drinking Water of
the Distribution System of Hyderabad City | 6,446 |
In most digital communication systems, bandwidth limited channel along with multipath propagation causes ISI (Inter Symbol Interference) to occur. This phenomenon causes distortion of the given transmitted symbol due to other transmitted symbols. With the help of equalization ISI can be reduced. This paper presents a solution to the ISI problem by performing blind equalization using ANN (Artificial Neural Networks). The simulated network is a multilayer feedforward Perceptron ANN, which has been trained by utilizing the error back-propagation algorithm. The weights of the network are updated in accordance with training of the network. This paper presents a very effective method for blind channel equalization, being more efficient than the pre-existing algorithms. The obtained results show a visible reduction in the noise content. | Channel Equalization Using Multilayer Perceptron Networks | 6,447 |
Heuristic algorithms are able to optimize objective functions efficiently because they use intelligently the information about the objective functions. Thus, information utilization is critical to the performance of heuristics. However, the concept of information utilization has remained vague and abstract because there is no reliable metric to reflect the extent to which the information about the objective function is utilized by heuristic algorithms. In this paper, the metric of information utilization ratio (IUR) is defined, which is the ratio of the utilized information quantity over the acquired information quantity in the search process. The IUR proves to be well-defined. Several examples of typical heuristic algorithms are given to demonstrate the procedure of calculating the IUR. Empirical evidences on the correlation between the IUR and the performance of a heuristic are also provided. The IUR can be an index of how finely an algorithm is designed and guide the invention of new heuristics and the improvement of existing ones. | Information Utilization Ratio in Heuristic Optimization Algorithms | 6,448 |
We propose a novel activation function that implements piece-wise orthogonal non-linear mappings based on permutations. It is straightforward to implement, and very computationally efficient, also it has little memory requirements. We tested it on two toy problems for feedforward and recurrent networks, it shows similar performance to tanh and ReLU. OPLU activation function ensures norm preservance of the backpropagated gradients, therefore it is potentially good for the training of deep, extra deep, and recurrent neural networks. | Norm-preserving Orthogonal Permutation Linear Unit Activation Functions
(OPLU) | 6,449 |
This paper introduces two recurrent neural network structures called Simple Gated Unit (SGU) and Deep Simple Gated Unit (DSGU), which are general structures for learning long term dependencies. Compared to traditional Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU), both structures require fewer parameters and less computation time in sequence classification tasks. Unlike GRU and LSTM, which require more than one gates to control information flow in the network, SGU and DSGU only use one multiplicative gate to control the flow of information. We show that this difference can accelerate the learning speed in tasks that require long dependency information. We also show that DSGU is more numerically stable than SGU. In addition, we also propose a standard way of representing inner structure of RNN called RNN Conventional Graph (RCG), which helps analyzing the relationship between input units and hidden units of RNN. | Deep Gate Recurrent Neural Network | 6,450 |
Neural networks and evolutionary computation have a rich intertwined history. They most commonly appear together when an evolutionary algorithm optimises the parameters and topology of a neural network for reinforcement learning problems, or when a neural network is applied as a surrogate fitness function to aid the evolutionary optimisation of expensive fitness functions. In this paper we take a different approach, asking the question of whether a neural network can be used to provide a mutation distribution for an evolutionary algorithm, and what advantages this approach may offer? Two modern neural network models are investigated, a Denoising Autoencoder modified to produce stochastic outputs and the Neural Autoregressive Distribution Estimator. Results show that the neural network approach to learning genotypes is able to solve many difficult discrete problems, such as MaxSat and HIFF, and regularly outperforms other evolutionary techniques. | Learning to Generate Genotypes with Neural Networks | 6,451 |
Volatility in stock markets has been extensively studied in the applied finance literature. In this paper, Artificial Neural Network models based on various back propagation algorithms have been constructed to predict volatility in the Indian stock market through volatility of NIFTY returns and volatility of gold returns. This model considers India VIX, CBOE VIX, volatility of crude oil returns (CRUDESDR), volatility of DJIA returns (DJIASDR), volatility of DAX returns (DAXSDR), volatility of Hang Seng returns (HANGSDR) and volatility of Nikkei returns (NIKKEISDR) as predictor variables. Three sets of experiments have been performed over three time periods to judge the effectiveness of the approach. | Forecasting Volatility in Indian Stock Market using Artificial Neural
Network with Multiple Inputs and Outputs | 6,452 |
In this article, a novel neuro-inspired low-resolution online unsupervised learning rule is proposed to train the reservoir or liquid of Liquid State Machine. The liquid is a sparsely interconnected huge recurrent network of spiking neurons. The proposed learning rule is inspired from structural plasticity and trains the liquid through formation and elimination of synaptic connections. Hence, the learning involves rewiring of the reservoir connections similar to structural plasticity observed in biological neural networks. The network connections can be stored as a connection matrix and updated in memory by using Address Event Representation (AER) protocols which are generally employed in neuromorphic systems. On investigating the 'pairwise separation property' we find that trained liquids provide 1.36 $\pm$ 0.18 times more inter-class separation while retaining similar intra-class separation as compared to random liquids. Moreover, analysis of the 'linear separation property' reveals that trained liquids are 2.05 $\pm$ 0.27 times better than random liquids. Furthermore, we show that our liquids are able to retain the 'generalization' ability and 'generality' of random liquids. A memory analysis shows that trained liquids have 83.67 $\pm$ 5.79 ms longer fading memory than random liquids which have shown 92.8 $\pm$ 5.03 ms fading memory for a particular type of spike train inputs. We also throw some light on the dynamics of the evolution of recurrent connections within the liquid. Moreover, compared to 'Separation Driven Synaptic Modification' - a recently proposed algorithm for iteratively refining reservoirs, our learning rule provides 9.30%, 15.21% and 12.52% more liquid separations and 2.8%, 9.1% and 7.9% better classification accuracies for four, eight and twelve class pattern recognition tasks respectively. | An Online Structural Plasticity Rule for Generating Better Reservoirs | 6,453 |
Modern films, games and virtual reality applications are dependent on convincing computer graphics. Highly complex models are a requirement for the successful delivery of many scenes and environments. While workflows such as rendering, compositing and animation have been streamlined to accommodate increasing demands, modelling complex models is still a laborious task. This paper introduces the computational benefits of an Interactive Genetic Algorithm (IGA) to computer graphics modelling while compensating the effects of user fatigue, a common issue with Interactive Evolutionary Computation. An intelligent agent is used in conjunction with an IGA that offers the potential to reduce the effects of user fatigue by learning from the choices made by the human designer and directing the search accordingly. This workflow accelerates the layout and distribution of basic elements to form complex models. It captures the designer's intent through interaction, and encourages playful discovery. | Multi-agent evolutionary systems for the generation of complex virtual
worlds | 6,454 |
Evolutionary algorithms have been widely studied from a theoretical perspective. In particular, the area of runtime analysis has contributed significantly to a theoretical understanding and provided insights into the working behaviour of these algorithms. We study how these insights into evolutionary processes can be used for evolutionary art. We introduce the notion of evolutionary image transition which transfers a given starting image into a target image through an evolutionary process. Combining standard mutation effects known from the optimization of the classical benchmark function OneMax and different variants of random walks, we present ways of performing evolutionary image transition with different artistic effects. | Evolutionary Image Transition Based on Theoretical Insights of Random
Processes | 6,455 |
There has been a variety of crossover operators proposed for Real-Coded Genetic Algorithms (RCGAs), which recombine values from the same location in pairs of strings. In this article we present a recombination operator for RC- GAs that selects the locations randomly in both parents, and compare it to mainstream crossover operators in a set of experiments on a range of standard multidimensional optimization problems and a clustering problem. We present two variants of the operator, either selecting both bits uniformly at random in the strings, or sampling the second bit from a normal distribution centered at the selected location in the first string. While the operator is biased towards exploitation of fitness space, the random selection of the second bit for swap- ping makes it slightly less exploitation-biased. Extensive statistical analysis using a non-parametric test shows the advantage of the new recombination operators on a range of test functions. | K-Bit-Swap: A New Operator For Real-Coded Evolutionary Algorithms | 6,456 |
Benchmarks and datasets have important role in evaluation of machine learning algorithms and neural network implementations. Traditional dataset for images such as MNIST is applied to evaluate efficiency of different training algorithms in neural networks. This demand is different in Spiking Neural Networks (SNN) as they require spiking inputs. It is widely believed, in the biological cortex the timing of spikes is irregular. Poisson distributions provide adequate descriptions of the irregularity in generating appropriate spikes. Here, we introduce a spike-based version of MNSIT (handwritten digits dataset),using Poisson distribution and show the Poissonian property of the generated streams. We introduce a new version of evt_MNIST which can be used for neural network evaluation. | evt_MNIST: A spike based version of traditional MNIST | 6,457 |
Recurrent neural networks (RNN) are simple dynamical systems whose computational power has been attributed to their short-term memory. Short-term memory of RNNs has been previously studied analytically only for the case of orthogonal networks, and only under annealed approximation, and uncorrelated input. Here for the first time, we present an exact solution to the memory capacity and the task-solving performance as a function of the structure of a given network instance, enabling direct determination of the function--structure relation in RNNs. We calculate the memory capacity for arbitrary networks with exponentially correlated input and further related it to the performance of the system on signal processing tasks in a supervised learning setup. We compute the expected error and the worst-case error bound as a function of the spectra of the network and the correlation structure of its inputs and outputs. Our results give an explanation for learning and generalization of task solving using short-term memory, which is crucial for building alternative computer architectures using physical phenomena based on the short-term memory principle. | Memory and Information Processing in Recurrent Neural Networks | 6,458 |
This work presents the application of the artificial neural networks, trained and structurally optimized by genetic algorithms, for modeling of crude distillation process at PKN ORLEN S.A. refinery. Models for the main fractionator distillation column products were developed using historical data. Quality of the fractions were predicted based on several chosen process variables. The performance of the model was validated using test data. Neural networks used in companion with genetic algorithms proved that they can accurately predict fractions quality shifts, reproducing the results of the standard laboratory analysis. Simple knowledge extraction method from neural network model built was also performed. Genetic algorithms can be successfully utilized in efficient training of large neural networks and finding their optimal structures. | Application of artificial neural networks and genetic algorithms for
crude fractional distillation process modeling | 6,459 |
This paper concerns applications of genetic algorithms and genetic programming to tasks for which it is difficult to find a representation that does not map to a highly complex and discontinuous fitness landscape. In such cases the standard algorithm is prone to getting trapped in local extremes. The paper proposes several adaptive mechanisms that are useful in preventing the search from getting trapped. | Fitness-based Adaptive Control of Parameters in Genetic Programming:
Adaptive Value Setting of Mutation Rate and Flood Mechanisms | 6,460 |
This document details the rationales behind assessing the performance of numerical black-box optimizers on multi-objective problems within the COCO platform and in particular on the biobjective test suite bbob-biobj. The evaluation is based on a hypervolume of all non-dominated solutions in the archive of candidate solutions and measures the runtime until the hypervolume value succeeds prescribed target values. | Biobjective Performance Assessment with the COCO Platform | 6,461 |
This research is focused on solving problems in the area of software project management using metaheuristic search algorithms and as such is research in the field of search based software engineering. The main aim of this research is to evaluate the performance of different metaheuristic search techniques in resource allocation and scheduling problems that would be typical of software development projects. This paper reports a set of experiments which evaluate the performance of three algorithms, namely simulated annealing, tabu search and genetic algorithms. The experimental results indicate that all of the metaheuristics search techniques can be used to solve problems in resource allocation and scheduling within a software project. Finally, a comparative analysis suggests that overall the genetic algorithm had performed better than simulated annealing and tabu search. | Resource allocation using metaheuristic search | 6,462 |
Previous RNN architectures have largely been superseded by LSTM, or "Long Short-Term Memory". Since its introduction, there have been many variations on this simple design. However, it is still widely used and we are not aware of a gated-RNN architecture that outperforms LSTM in a broad sense while still being as simple and efficient. In this paper we propose a modified LSTM-like architecture. Our architecture is still simple and achieves better performance on the tasks that we tested on. We also introduce a new RNN performance benchmark that uses the handwritten digits and stresses several important network capabilities. | LSTM with Working Memory | 6,463 |
We present four training and prediction schedules from the same character-level recurrent neural network. The efficiency of these schedules is tested in terms of model effectiveness as a function of training time and amount of training data seen. We show that the choice of training and prediction schedule potentially has a considerable impact on the prediction effectiveness for a given training budget. | Efficiency Evaluation of Character-level RNN Training Schedules | 6,464 |
We propose a multi-objective optimization algorithm aimed at achieving good anytime performance over a wide range of problems. Performance is assessed in terms of the hypervolume metric. The algorithm called HMO-CMA-ES represents a hybrid of several old and new variants of CMA-ES, complemented by BOBYQA as a warm start. We benchmark HMO-CMA-ES on the recently introduced bi-objective problem suite of the COCO framework (COmparing Continuous Optimizers), consisting of 55 scalable continuous optimization problems, which is used by the Black-Box Optimization Benchmarking (BBOB) Workshop 2016. | Anytime Bi-Objective Optimization with a Hybrid Multi-Objective CMA-ES
(HMO-CMA-ES) | 6,465 |
We present an any-time performance assessment for benchmarking numerical optimization algorithms in a black-box scenario, applied within the COCO benchmarking platform. The performance assessment is based on runtimes measured in number of objective function evaluations to reach one or several quality indicator target values. We argue that runtime is the only available measure with a generic, meaningful, and quantitative interpretation. We discuss the choice of the target values, runlength-based targets, and the aggregation of results by using simulated restarts, averages, and empirical distribution functions. | COCO: Performance Assessment | 6,466 |
This paper is dedicated to the long-term, or multi-step-ahead, time series prediction problem. We propose a novel method for training feed-forward neural networks, such as multilayer perceptrons, with tapped delay lines. Special batch calculation of derivatives called Forecasted Propagation Through Time and batch modification of the Extended Kalman Filter are introduced. Experiments were carried out on well-known time series benchmarks, the Mackey-Glass chaotic process and the Santa Fe Laser Data Series. Recurrent and feed-forward neural networks were evaluated. | Direct Method for Training Feed-forward Neural Networks using Batch
Extended Kalman Filter for Multi-Step-Ahead Predictions | 6,467 |
In the neuroevolution literature, research has primarily focused on evolving the number of nodes, connections, and weights in artificial neural networks. Few attempts have been made to evolve activation functions. Research in evolving activation functions has mainly focused on evolving function parameters, and developing heterogeneous networks by selecting from a fixed pool of activation functions. This paper introduces a novel technique for evolving heterogeneous artificial neural networks through combinatorially generating piecewise activation functions to enhance expressive power. I demonstrate this technique on NeuroEvolution of Augmenting Topologies using ArcTan and Sigmoid, and show that it outperforms the original algorithm on non-Markovian double pole balancing. This technique expands the landscape of unconventional activation functions by demonstrating that they are competitive with canonical choices, and introduces a purview for further exploration of automatic model selection for artificial neural networks. | Combinatorially Generated Piecewise Activation Functions | 6,468 |
A co-evolutionary algorithm (CA) based chess player is presented. Implementation details of the algorithms, namely coding, population, variation operators are described. The alpha-beta or mini-max like behaviour of the player is achieved through two competitive or cooperative populations. Special attention is given to the fitness function evaluation (the heart of the solution). Test results on algorithms vs. algorithms or human player is provided. | Chess Player by Co-Evolutionary Algorithm | 6,469 |
Most of the problems in genetic algorithms are very complex and demand a large amount of resources that current technology can not offer. Our purpose was to develop a Java-JINI distributed library that implements Genetic Algorithms with sub-populations (coarse grain) and a graphical interface in order to configure and follow the evolution of the search. The sub-populations are simulated/evaluated in personal computers connected trough a network, keeping in mind different models of sub-populations, migration policies and network topologies. We show that this model delays the convergence of the population keeping a higher level of genetic diversity and allows a much greater number of evaluations since they are distributed among several computers compared with the traditional Genetic Algorithms. | Evolutionary Demographic Algorithms | 6,470 |
Deep Neural Networks (DNN) have achieved human level performance in many image analytics tasks but DNNs are mostly deployed to GPU platforms that consume a considerable amount of power. Brain-inspired spiking neuromorphic chips consume low power and can be highly parallelized. However, for deploying DNNs to energy efficient neuromorphic chips the incompatibility between continuous neurons and synaptic weights of traditional DNNs, discrete spiking neurons and synapses of neuromorphic chips has to be overcome. Previous work has achieved this by training a network to learn continuous probabilities and deployment to a neuromorphic architecture by random sampling these probabilities. An ensemble of sampled networks is needed to approximate the performance of the trained network. In the work presented in this paper, we have extended previous research by directly learning binary synaptic crossbars. Results on MNIST show that better performance can be achieved with a small network in one time step (92.7% maximum observed accuracy vs 95.98% accuracy in our work). Top results on a larger network are similar to previously published results (99.42% maximum observed accuracy vs 99.45% accuracy in our work). More importantly, in our work a smaller ensemble is needed to achieve similar or better accuracy than previous work, which translates into significantly decreased energy consumption for both networks. Results of our work are stable since they do not require random sampling. | Improving energy efficiency and classification accuracy of neuromorphic
chips by learning binary synaptic crossbars | 6,471 |
The performance of different mutation operators is usually evaluated in conjunc-tion with specific parameter settings of genetic algorithms and target problems. Most studies focus on the classical genetic algorithm with different parameters or on solving unconstrained combinatorial optimization problems such as the traveling salesman problems. In this paper, a subpopulation-based genetic al-gorithm that uses only mutation and selection is developed to solve multi-robot task allocation problems. The target problems are constrained combinatorial optimization problems, and are more complex if cooperative tasks are involved as these introduce additional spatial and temporal constraints. The proposed genetic algorithm can obtain better solutions than classical genetic algorithms with tournament selection and partially mapped crossover. The performance of different mutation operators in solving problems without/with cooperative tasks is evaluated. The results imply that inversion mutation performs better than others when solving problems without cooperative tasks, and the swap-inversion combination performs better than others when solving problems with cooperative tasks. | On the performance of different mutation operators of a
subpopulation-based genetic algorithm for multi-robot task allocation
problems | 6,472 |
Spiking neural networks (SNNs) with adaptive synapses reflect core properties of biological neural networks. Speech recognition, as an application involving audio coding and dynamic learning, provides a good test problem to study SNN functionality. We present a simple, novel, and efficient nonrecurrent SNN that learns to convert a speech signal into a spike train signature. The signature is distinguishable from signatures for other speech signals representing different words, thereby enabling digit recognition and discrimination in devices that use only spiking neurons. The method uses a small, nonrecurrent SNN consisting of Izhikevich neurons equipped with spike timing dependent plasticity (STDP) and biologically realistic synapses. This approach introduces an efficient and fast network without error-feedback training, although it does require supervised training. The new simulation results produce discriminative spike train patterns for spoken digits in which highly correlated spike trains belong to the same category and low correlated patterns belong to different categories. The proposed SNN is evaluated using a spoken digit recognition task where a subset of the Aurora speech dataset is used. The experimental results show that the network performs well in terms of accuracy rate and complexity. | A Spiking Network that Learns to Extract Spike Signatures from Speech
Signals | 6,473 |
It is of some interest to understand how statistically based mechanisms for signal processing might be integrated with biologically motivated mechanisms such as neural networks. This paper explores a novel hybrid approach for classifying segments of sequential data, such as individual spoken works. The approach combines a hidden Markov model (HMM) with a spiking neural network (SNN). The HMM, consisting of states and transitions, forms a fixed backbone with nonadaptive transition probabilities. The SNN, however, implements a biologically based Bayesian computation that derives from the spike timing-dependent plasticity (STDP) learning rule. The emission (observation) probabilities of the HMM are represented in the SNN and trained with the STDP rule. A separate SNN, each with the same architecture, is associated with each of the states of the HMM. Because of the STDP training, each SNN implements an expectation maximization algorithm to learn the emission probabilities for one HMM state. The model was studied on synthesized spike-train data and also on spoken word data. Preliminary results suggest its performance compares favorably with other biologically motivated approaches. Because of the model's uniqueness and initial promise, it warrants further study. It provides some new ideas on how the brain might implement the equivalent of an HMM in a neural circuit. | Training a Hidden Markov Model with a Bayesian Spiking Neural Network | 6,474 |
The final version of this paper has been published in IEEEXplore available at http://ieeexplore.ieee.org/document/7727213. Please cite this paper as: Amirhossein Tavanaei, Timothee Masquelier, and Anthony Maida, Acquisition of visual features through probabilistic spike-timing-dependent plasticity. IEEE International Joint Conference on Neural Networks. pp. 307-314, IJCNN 2016. This paper explores modifications to a feedforward five-layer spiking convolutional network (SCN) of the ventral visual stream [Masquelier, T., Thorpe, S., Unsupervised learning of visual features through spike timing dependent plasticity. PLoS Computational Biology, 3(2), 247-257]. The original model showed that a spike-timing-dependent plasticity (STDP) learning algorithm embedded in an appropriately selected SCN could perform unsupervised feature discovery. The discovered features where interpretable and could effectively be used to perform rapid binary decisions in a classifier. In order to study the robustness of the previous results, the present research examines the effects of modifying some of the components of the original model. For improved biological realism, we replace the original non-leaky integrate-and-fire neurons with Izhikevich-like neurons. We also replace the original STDP rule with a novel rule that has a probabilistic interpretation. The probabilistic STDP slightly but significantly improves the performance for both types of model neurons. Use of the Izhikevich-like neuron was not found to improve performance although performance was still comparable to the IF neuron. This shows that the model is robust enough to handle more biologically realistic neurons. We also conclude that the underlying reasons for stable performance in the model are preserved despite the overt changes to the explicit components of the model. | Acquisition of Visual Features Through Probabilistic
Spike-Timing-Dependent Plasticity | 6,475 |
The paper investigates a new type of truly critical echo state networks where individual transfer functions for every neuron can be modified to anticipate the expected next input. Deviations from expected input are only forgotten slowly in power law fashion. The paper outlines the theory, numerically analyzes a one neuron model network and finally discusses technical and also biological implications of this type of approach. | Critical Echo State Networks that Anticipate Input using Morphable
Transfer Functions | 6,476 |
The article, after a brief introduction on genetic algorithms and their functioning, presents a kind of genetic algorithm called Viral Search. We present the key concepts, we formally derive the algorithm and we perform numerical tests designed to illustrate the potential and limits. | Viral Search algorithm | 6,477 |
The fields of neural computation and artificial neural networks have developed much in the last decades. Most of the works in these fields focus on implementing and/or learning discrete functions or behavior. However, technical, physical, and also cognitive processes evolve continuously in time. This cannot be described directly with standard architectures of artificial neural networks such as multi-layer feed-forward perceptrons. Therefore, in this paper, we will argue that neural networks modeling continuous time are needed explicitly for this purpose, because with them the synthesis and analysis of continuous and possibly periodic processes in time are possible (e.g. for robot behavior) besides computing discrete classification functions (e.g. for logical reasoning). We will relate possible neural network architectures with (hybrid) automata models that allow to express continuous processes. | Neural Networks and Continuous Time | 6,478 |
Evolutionary algorithms (EAs) have been well acknowledged as a promising paradigm for solving optimisation problems with multiple conflicting objectives in the sense that they are able to locate a set of diverse approximations of Pareto optimal solutions in a single run. EAs drive the search for approximated solutions through maintaining a diverse population of solutions and by recombining promising solutions selected from the population. Combining machine learning techniques has shown great potentials since the intrinsic structure of the Pareto optimal solutions of an multiobjective optimisation problem can be learned and used to guide for effective recombination. However, existing multiobjective EAs (MOEAs) based on structure learning spend too much computational resources on learning. To address this problem, we propose to use an online learning scheme. Based on the fact that offsprings along evolution are streamy, dependent and non-stationary (which implies that the intrinsic structure, if any, is temporal and scale-variant), an online agglomerative clustering algorithm is applied to adaptively discover the intrinsic structure of the Pareto optimal solution set; and to guide effective offspring recombination. Experimental results have shown significant improvement over five state-of-the-art MOEAs on a set of well-known benchmark problems with complicated Pareto sets and complex Pareto fronts. | Learning from Non-Stationary Stream Data in Multiobjective Evolutionary
Algorithm | 6,479 |
The paper is devoted to upper bounds on the expected first hitting times of the sets of local or global optima for non-elitist genetic algorithms with very high selection pressure. The results of this paper extend the range of situations where the upper bounds on the expected runtime are known for genetic algorithms and apply, in particular, to the Canonical Genetic Algorithm. The obtained bounds do not require the probability of fitness-decreasing mutation to be bounded by a constant less than one. | Hitting times of local and global optima in genetic algorithms with very
high selection pressure | 6,480 |
While gradient descent has proven highly successful in learning connection weights for neural networks, the actual structure of these networks is usually determined by hand, or by other optimization algorithms. Here we describe a simple method to make network structure differentiable, and therefore accessible to gradient descent. We test this method on recurrent neural networks applied to simple sequence prediction problems. Starting with initial networks containing only one node, the method automatically builds networks that successfully solve the tasks. The number of nodes in the final network correlates with task difficulty. The method can dynamically increase network size in response to an abrupt complexification in the task; however, reduction in network size in response to task simplification is not evident for reasonable meta-parameters. The method does not penalize network performance for these test tasks: variable-size networks actually reach better performance than fixed-size networks of higher, lower or identical size. We conclude by discussing how this method could be applied to more complex networks, such as feedforward layered networks, or multiple-area networks of arbitrary shape. | Neural networks with differentiable structure | 6,481 |
Over the past 30 years many researchers in the field of evolutionary computation have put a lot of effort to introduce various approaches for solving hard problems. Most of these problems have been inspired by major industries so that solving them, by providing either optimal or near optimal solution, was of major significance. Indeed, this was a very promising trajectory as advances in these problem-solving approaches could result in adding values to major industries. In this paper we revisit this trajectory to find out whether the attempts that started three decades ago are still aligned with the same goal, as complexities of real-world problems increased significantly. We present some examples of modern real-world problems, discuss why they might be difficult to solve, and whether there is any mismatch between these examples and the problems that are investigated in the evolutionary computation area. | Evolutionary computation for multicomponent problems: opportunities and
future directions | 6,482 |
In this paper we present a modified version of the Hyperbolic Tangent Activation Function as a learning unit generator for neural networks. The function uses an integer calibration constant as an approximation to the Euler number, e, based on a quadratic Real Number Formula (RNF) algorithm and an adaptive normalization constraint on the input activations to avoid the vanishing gradient. We demonstrate the effectiveness of the proposed modification using a hypothetical and real world dataset and show that lower run-times can be achieved by learning algorithms using this function leading to improved speed-ups and learning accuracies during training. | A Modified Activation Function with Improved Run-Times For Neural
Networks | 6,483 |
Computer simulations have become a very powerful tool for scientific research. Given the vast complexity that comes with many open scientific questions, a purely analytical or experimental approach is often not viable. For example, biological systems (such as the human brain) comprise an extremely complex organization and heterogeneous interactions across different spatial and temporal scales. In order to facilitate research on such problems, the BioDynaMo project (\url{https://biodynamo.web.cern.ch/}) aims at a general platform for computer simulations for biological research. Since the scientific investigations require extensive computer resources, this platform should be executable on hybrid cloud computing systems, allowing for the efficient use of state-of-the-art computing technology. This paper describes challenges during the early stages of the software development process. In particular, we describe issues regarding the implementation and the highly interdisciplinary as well as international nature of the collaboration. Moreover, we explain the methodologies, the approach, and the lessons learnt by the team during these first stages. | The BioDynaMo Project | 6,484 |
Spike-timing-dependent plasticity (STDP) incurs both causal and acausal synaptic weight updates, for negative and positive time differences between pre-synaptic and post-synaptic spike events. For realizing such updates in neuromorphic hardware, current implementations either require forward and reverse lookup access to the synaptic connectivity table, or rely on memory-intensive architectures such as crossbar arrays. We present a novel method for realizing both causal and acausal weight updates using only forward lookup access of the synaptic connectivity table, permitting memory-efficient implementation. A simplified implementation in FPGA, using a single timer variable for each neuron, closely approximates exact STDP cumulative weight updates for neuron refractory periods greater than 10 ms, and reduces to exact STDP for refractory periods greater than the STDP time window. Compared to conventional crossbar implementation, the forward table-based implementation leads to substantial memory savings for sparsely connected networks supporting scalable neuromorphic systems with fully reconfigurable synaptic connectivity and plasticity. | Forward Table-Based Presynaptic Event-Triggered Spike-Timing-Dependent
Plasticity | 6,485 |
We provide a rigorous runtime analysis concerning the update strength, a vital parameter in probabilistic model-building GAs such as the step size $1/K$ in the compact Genetic Algorithm (cGA) and the evaporation factor $\rho$ in ACO. While a large update strength is desirable for exploitation, there is a general trade-off: too strong updates can lead to genetic drift and poor performance. We demonstrate this trade-off for the cGA and a simple MMAS ACO algorithm on the OneMax function. More precisely, we obtain lower bounds on the expected runtime of $\Omega(K\sqrt{n} + n \log n)$ and $\Omega(\sqrt{n}/\rho + n \log n)$, respectively, showing that the update strength should be limited to $1/K, \rho = O(1/(\sqrt{n} \log n))$. In fact, choosing $1/K, \rho \sim 1/(\sqrt{n}\log n)$ both algorithms efficiently optimize OneMax in expected time $O(n \log n)$. Our analyses provide new insights into the stochastic behavior of probabilistic model-building GAs and propose new guidelines for setting the update strength in global optimization. | Update Strength in EDAs and ACO: How to Avoid Genetic Drift | 6,486 |
This paper introduces an Enhanced Boolean version of the Correlation Matrix Memory (CMM), which is useful to work with binary memories. A novel Boolean Orthonormalization Process (BOP) is presented to convert a non-orthonormal Boolean basis, i.e., a set of non-orthonormal binary vectors (in a Boolean sense) to an orthonormal Boolean basis, i.e., a set of orthonormal binary vectors (in a Boolean sense). This work shows that it is possible to improve the performance of Boolean CMM thanks BOP algorithm. Besides, the BOP algorithm has a lot of additional fields of applications, e.g.: Steganography, Hopfield Networks, Bi-level image processing, etc. Finally, it is important to mention that the BOP is an extremely stable and fast algorithm. | Enhanced Boolean Correlation Matrix Memory | 6,487 |
Multi-population evolutionary algorithms are, by nature, highly complex and difficult to describe. Even two populations working in concert (or opposition) present a myriad of potential configurations that are often difficult to relate using text alone. Little effort has been made, however, to depict these kinds of systems, relying solely on the simple structural connections (related using ad hoc diagrams) between populations and often leaving out crucial details. In this paper, we propose a notation and accompanying formalism for consistently and powerfully depicting these structures and the relationships within them in an intuitive and consistent way. Using our notation, we examine simple co-evolutionary systems and discover new configurations by the simple process of "drawing on a whiteboard". Finally, we demonstrate that even complex, highly-interconnected systems with large numbers of populations can be understood with ease using the advanced features of our formalism | mpEAd: Multi-Population EA Diagrams | 6,488 |
Evolutionary multitasking has recently emerged as a novel paradigm that enables the similarities and/or latent complementarities (if present) between distinct optimization tasks to be exploited in an autonomous manner simply by solving them together with a unified solution representation scheme. An important matter underpinning future algorithmic advancements is to develop a better understanding of the driving force behind successful multitask problem-solving. In this regard, two (seemingly disparate) ideas have been put forward, namely, (a) implicit genetic transfer as the key ingredient facilitating the exchange of high-quality genetic material across tasks, and (b) population diversification resulting in effective global search of the unified search space encompassing all tasks. In this paper, we present some empirical results that provide a clearer picture of the relationship between the two aforementioned propositions. For the numerical experiments we make use of Sudoku puzzles as case studies, mainly because of their feature that outwardly unlike puzzle statements can often have nearly identical final solutions. The experiments reveal that while on many occasions genetic transfer and population diversity may be viewed as two sides of the same coin, the wider implication of genetic transfer, as shall be shown herein, captures the true essence of evolutionary multitasking to the fullest. | Genetic Transfer or Population Diversification? Deciphering the Secret
Ingredients of Evolutionary Multitask Optimization | 6,489 |
The ability to backpropagate stochastic gradients through continuous latent distributions has been crucial to the emergence of variational autoencoders and stochastic gradient variational Bayes. The key ingredient is an unbiased and low-variance way of estimating gradients with respect to distribution parameters from gradients evaluated at distribution samples. The "reparameterization trick" provides a class of transforms yielding such estimators for many continuous distributions, including the Gaussian and other members of the location-scale family. However the trick does not readily extend to mixture density models, due to the difficulty of reparameterizing the discrete distribution over mixture weights. This report describes an alternative transform, applicable to any continuous multivariate distribution with a differentiable density function from which samples can be drawn, and uses it to derive an unbiased estimator for mixture density weight derivatives. Combined with the reparameterization trick applied to the individual mixture components, this estimator makes it straightforward to train variational autoencoders with mixture-distributed latent variables, or to perform stochastic variational inference with a mixture density variational posterior. | Stochastic Backpropagation through Mixture Density Distributions | 6,490 |
We propose a geometric model-free causality measurebased on multivariate delay embedding that can efficiently detect linear and nonlinear causal interactions between time series with no prior information. We then exploit the proposed causal interaction measure in real MEG data analysis. The results are used to construct effective connectivity maps of brain activity to decode different categories of visual stimuli. Moreover, we discovered that the MEG-based effective connectivity maps as a response to structured images exhibit more geometric patterns, as disclosed by analyzing the evolution of toplogical structures of the underlying networks using persistent homology. Extensive simulation and experimental result have been carried out to substantiate the capabilities of the proposed approach. | Effective Connectivity-Based Neural Decoding: A Causal
Interaction-Driven Approach | 6,491 |
PSO is a widely recognized optimization algorithm inspired by social swarm. In this brief we present a heterogeneous strategy particle swarm optimization (HSPSO), in which a proportion of particles adopt a fully informed strategy to enhance the converging speed while the rest are singly informed to maintain the diversity. Our extensive numerical experiments show that HSPSO algorithm is able to obtain satisfactory solutions, outperforming both PSO and the fully informed PSO. The evolution process is examined from both structural and microscopic points of view. We find that the cooperation between two types of particles can facilitate a good balance between exploration and exploitation, yielding better performance. We demonstrate the applicability of HSPSO on the filter design problem. | Heterogeneous Strategy Particle Swarm Optimization | 6,492 |
Evolutionary algorithms have been used in many ways to generate digital art. We study how evolutionary processes are used for evolutionary art and present a new approach to the transition of images. Our main idea is to define evolutionary processes for digital image transition, combining different variants of mutation and evolutionary mechanisms. We introduce box and strip mutation operators which are specifically designed for image transition. Our experimental results show that the process of an evolutionary algorithm in combination with these mutation operators can be used as a valuable way to produce unique generative art. | The Evolutionary Process of Image Transition in Conjunction with Box and
Strip Mutation | 6,493 |
This paper is a brief update on developments in the BioDynaMo project, a new platform for computer simulations for biological research. We will discuss the new capabilities of the simulator, important new concepts simulation methodology as well as its numerous applications to the computational biology and nanoscience communities. | The BioDynaMo Project: a platform for computer simulations of biological
dynamics | 6,494 |
In this paper, we are concerned with a branch of evolutionary algorithms termed estimation of distribution (EDA), which has been successfully used to tackle derivative-free global optimization problems. For existent EDA algorithms, it is a common practice to use a Gaussian distribution or a mixture of Gaussian components to represent the statistical property of available promising solutions found so far. Observing that the Student's t distribution has heavier and longer tails than the Gaussian, which may be beneficial for exploring the solution space, we propose a novel EDA algorithm termed ESTDA, in which the Student's t distribution, rather than Gaussian, is employed. To address hard multimodal and deceptive problems, we extend ESTDA further by substituting a single Student's t distribution with a mixture of Student's t distributions. The resulting algorithm is named as estimation of mixture of Student's t distribution algorithm (EMSTDA). Both ESTDA and EMSTDA are evaluated through extensive and in-depth numerical experiments using over a dozen of benchmark objective functions. Empirical results demonstrate that the proposed algorithms provide remarkably better performance than their Gaussian counterparts. | Student's t Distribution based Estimation of Distribution Algorithms for
Derivative-free Global Optimization | 6,495 |
Existing studies on dynamic multi-objective optimization focus on problems with time-dependent objective functions, while the ones with a changing number of objectives have rarely been considered in the literature. Instead of changing the shape or position of the Pareto-optimal front/set when having time-dependent objective functions, increasing or decreasing the number of objectives usually leads to the expansion or contraction of the dimension of the Pareto-optimal front/set manifold. Unfortunately, most existing dynamic handling techniques can hardly be adapted to this type of dynamics. In this paper, we report our attempt toward tackling the dynamic multi-objective optimization problems with a changing number of objectives. We implement a new two-archive evolutionary algorithm which maintains two co-evolving populations simultaneously. In particular, these two populations are complementary to each other: one concerns more about the convergence while the other concerns more about the diversity. The compositions of these two populations are adaptively reconstructed once the environment changes. In addition, these two populations interact with each other via a mating selection mechanism. Comprehensive experiments are conducted on various benchmark problems with a time-dependent number of objectives. Empirical results fully demonstrate the effectiveness of our proposed algorithm. | Dynamic Multi-Objectives Optimization with a Changing Number of
Objectives | 6,496 |
Recurrent Neural Networks (RNNs) produce state-of-art performance on many machine learning tasks but their demand on resources in terms of memory and computational power are often high. Therefore, there is a great interest in optimizing the computations performed with these models especially when considering development of specialized low-power hardware for deep networks. One way of reducing the computational needs is to limit the numerical precision of the network weights and biases. This has led to different proposed rounding methods which have been applied so far to only Convolutional Neural Networks and Fully-Connected Networks. This paper addresses the question of how to best reduce weight precision during training in the case of RNNs. We present results from the use of different stochastic and deterministic reduced precision training methods applied to three major RNN types which are then tested on several datasets. The results show that the weight binarization methods do not work with the RNNs. However, the stochastic and deterministic ternarization, and pow2-ternarization methods gave rise to low-precision RNNs that produce similar and even higher accuracy on certain datasets therefore providing a path towards training more efficient implementations of RNNs in specialized hardware. | Recurrent Neural Networks With Limited Numerical Precision | 6,497 |
Recurrent Bistable Gradient Networks are attractor based neural networks characterized by bistable dynamics of each single neuron. Coupled together using linear interaction determined by the interconnection weights, these networks do not suffer from spurious states or very limited capacity anymore. Vladimir Chinarov and Michael Menzinger, who invented these networks, trained them using Hebb's learning rule. We show, that this way of computing the weights leads to unwanted behaviour and limitations of the networks capabilities. Furthermore we evince, that using the first order of Hintons Contrastive Divergence algorithm leads to a quite promising recurrent neural network. These findings are tested by learning images of the MNIST database for handwritten numbers. | About Learning in Recurrent Bistable Gradient Networks | 6,498 |
The balance between convergence and diversity is a key issue of evolutionary multi-objective optimization. The recently proposed stable matching-based selection provides a new perspective to handle this balance under the framework of decomposition multi-objective optimization. In particular, the stable matching between subproblems and solutions, which achieves an equilibrium between their mutual preferences, implicitly strikes a balance between the convergence and diversity. Nevertheless, the original stable matching model has a high risk of matching a solution with a unfavorable subproblem which finally leads to an imbalanced selection result. In this paper, we propose an adaptive two-level stable matching-based selection for decomposition multi-objective optimization. Specifically, borrowing the idea of stable matching with incomplete lists, we match each solution with one of its favorite subproblems by restricting the length of its preference list during the first-level stable matching. During the second-level stable matching, the remaining subproblems are thereafter matched with their favorite solutions according to the classic stable matching model. In particular, we develop an adaptive mechanism to automatically set the length of preference list for each solution according to its local competitiveness. The performance of our proposed method is validated and compared with several state-of-the-art evolutionary multi-objective optimization algorithms on 62 benchmark problem instances. Empirical results fully demonstrate the competitive performance of our proposed method on problems with complicated Pareto sets and those with more than three objectives. | Matching-Based Selection with Incomplete Lists for Decomposition
Multi-Objective Optimization | 6,499 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.