text
stringlengths 17
3.36M
| source
stringlengths 3
333
| __index_level_0__
int64 0
518k
|
|---|---|---|
This work considers the problem of estimating the parameters of negative mixture models, i.e. mixture models that possibly involve negative weights. The contributions of this paper are as follows. (i) We show that every rational probability distributions on strings, a representation which occurs naturally in spectral learning, can be computed by a negative mixture of at most two probabilistic automata (or HMMs). (ii) We propose a method to estimate the parameters of negative mixture models having a specific tensor structure in their low order observable moments. Building upon a recent paper on tensor decompositions for learning latent variable models, we extend this work to the broader setting of tensors having a symmetric decomposition with positive and negative weights. We introduce a generalization of the tensor power method for complex valued tensors, and establish theoretical convergence guarantees. (iii) We show how our approach applies to negative Gaussian mixture models, for which we provide some experiments.
|
Learning Negative Mixture Models by Tensor Decompositions
| 2,600
|
Motivated by multi-distribution divergences, which originate in information theory, we propose a notion of `multi-point' kernels, and study their applications. We study a class of kernels based on Jensen type divergences and show that these can be extended to measure similarity among multiple points. We study tensor flattening methods and develop a multi-point (kernel) spectral clustering (MSC) method. We further emphasize on a special case of the proposed kernels, which is a multi-point extension of the linear (dot-product) kernel and show the existence of cubic time tensor flattening algorithm in this case. Finally, we illustrate the usefulness of our contributions using standard data sets and image segmentation tasks.
|
Spectral Clustering with Jensen-type kernels and their multi-point
extensions
| 2,601
|
We tackle the problem of learning linear classifiers from noisy datasets in a multiclass setting. The two-class version of this problem was studied a few years ago by, e.g. Bylander (1994) and Blum et al. (1996): in these contributions, the proposed approaches to fight the noise revolve around a Perceptron learning scheme fed with peculiar examples computed through a weighted average of points from the noisy training set. We propose to build upon these approaches and we introduce a new algorithm called UMA (for Unconfused Multiclass additive Algorithm) which may be seen as a generalization to the multiclass setting of the previous approaches. In order to characterize the noise we use the confusion matrix as a multiclass extension of the classification noise studied in the aforementioned literature. Theoretically well-founded, UMA furthermore displays very good empirical noise robustness, as evidenced by numerical simulations conducted on both synthetic and real data. Keywords: Multiclass classification, Perceptron, Noisy labels, Confusion Matrix
|
Unconfused Ultraconservative Multiclass Algorithms
| 2,602
|
In many online learning problems we are interested in predicting local information about some universe of items. For example, we may want to know whether two items are in the same cluster rather than computing an assignment of items to clusters; we may want to know which of two teams will win a game rather than computing a ranking of teams. Although finding the optimal clustering or ranking is typically intractable, it may be possible to predict the relationships between items as well as if you could solve the global optimization problem exactly. Formally, we consider an online learning problem in which a learner repeatedly guesses a pair of labels (l(x), l(y)) and receives an adversarial payoff depending on those labels. The learner's goal is to receive a payoff nearly as good as the best fixed labeling of the items. We show that a simple algorithm based on semidefinite programming can obtain asymptotically optimal regret in the case where the number of possible labels is O(1), resolving an open problem posed by Hazan, Kale, and Shalev-Schwartz. Our main technical contribution is a novel use and analysis of the log determinant regularizer, exploiting the observation that log det(A + I) upper bounds the entropy of any distribution with covariance matrix A.
|
Online Local Learning via Semidefinite Programming
| 2,603
|
We provide an information-theoretic analysis of Thompson sampling that applies across a broad range of online optimization problems in which a decision-maker must learn from partial feedback. This analysis inherits the simplicity and elegance of information theory and leads to regret bounds that scale with the entropy of the optimal-action distribution. This strengthens preexisting results and yields new insight into how information improves performance.
|
An Information-Theoretic Analysis of Thompson Sampling
| 2,604
|
We propose information-directed sampling -- a new approach to online optimization problems in which a decision-maker must balance between exploration and exploitation while learning from partial feedback. Each action is sampled in a manner that minimizes the ratio between squared expected single-period regret and a measure of information gain: the mutual information between the optimal action and the next observation. We establish an expected regret bound for information-directed sampling that applies across a very general class of models and scales with the entropy of the optimal action distribution. We illustrate through simple analytic examples how information-directed sampling accounts for kinds of information that alternative approaches do not adequately address and that this can lead to dramatic performance gains. For the widely studied Bernoulli, Gaussian, and linear bandit problems, we demonstrate state-of-the-art simulation performance.
|
Learning to Optimize via Information-Directed Sampling
| 2,605
|
This paper revisits the problem of learning a k-CNF Boolean function from examples in the context of online learning under the logarithmic loss. In doing so, we give a Bayesian interpretation to one of Valiant's celebrated PAC learning algorithms, which we then build upon to derive two efficient, online, probabilistic, supervised learning algorithms for predicting the output of an unknown k-CNF Boolean function. We analyze the loss of our methods, and show that the cumulative log-loss can be upper bounded, ignoring logarithmic factors, by a polynomial function of the size of each example.
|
Online Learning of k-CNF Boolean Functions
| 2,606
|
This work investigates into cost behaviors of binary classification measures in a background of class-imbalanced problems. Twelve performance measures are studied, such as F measure, G-means in terms of accuracy rates, and of recall and precision, balance error rate (BER), Matthews correlation coefficient (MCC), Kappa coefficient, etc. A new perspective is presented for those measures by revealing their cost functions with respect to the class imbalance ratio. Basically, they are described by four types of cost functions. The functions provides a theoretical understanding why some measures are suitable for dealing with class-imbalanced problems. Based on their cost functions, we are able to conclude that G-means of accuracy rates and BER are suitable measures because they show "proper" cost behaviors in terms of "a misclassification from a small class will cause a greater cost than that from a large class". On the contrary, F1 measure, G-means of recall and precision, MCC and Kappa coefficient measures do not produce such behaviors so that they are unsuitable to serve our goal in dealing with the problems properly.
|
A study on cost behaviors of binary classification measures in
class-imbalanced problems
| 2,607
|
This paper presents an approximate method for performing Bayesian inference in models with conditional independence over a decentralized network of learning agents. The method first employs variational inference on each individual learning agent to generate a local approximate posterior, the agents transmit their local posteriors to other agents in the network, and finally each agent combines its set of received local posteriors. The key insight in this work is that, for many Bayesian models, approximate inference schemes destroy symmetry and dependencies in the model that are crucial to the correct application of Bayes' rule when combining the local posteriors. The proposed method addresses this issue by including an additional optimization step in the combination procedure that accounts for these broken dependencies. Experiments on synthetic and real data demonstrate that the decentralized method provides advantages in computational performance and predictive test likelihood over previous batch and distributed methods.
|
Approximate Decentralized Bayesian Inference
| 2,608
|
Many kernel methods suffer from high time and space complexities and are thus prohibitive in big-data applications. To tackle the computational challenge, the Nystr\"om method has been extensively used to reduce time and space complexities by sacrificing some accuracy. The Nystr\"om method speedups computation by constructing an approximation of the kernel matrix using only a few columns of the matrix. Recently, a variant of the Nystr\"om method called the modified Nystr\"om method has demonstrated significant improvement over the standard Nystr\"om method in approximation accuracy, both theoretically and empirically. In this paper, we propose two algorithms that make the modified Nystr\"om method practical. First, we devise a simple column selection algorithm with a provable error bound. Our algorithm is more efficient and easier to implement than and nearly as accurate as the state-of-the-art algorithm. Second, with the selected columns at hand, we propose an algorithm that computes the approximation in lower time complexity than the approach in the previous work. Furthermore, we prove that the modified Nystr\"om method is exact under certain conditions, and we establish a lower error bound for the modified Nystr\"om method.
|
Efficient Algorithms and Error Analysis for the Modified Nystrom Method
| 2,609
|
In this paper, we present a computational technique to deal with uncertainty in dynamic continuous models in Social Sciences. Considering data from surveys, the method consists of determining the probability distribution of the survey output and this allows to sample data and fit the model to the sampled data using a goodness-of-fit criterion based on the chi-square-test. Taking the fitted parameters non-rejected by the chi-square-test, substituting them into the model and computing their outputs, we build 95% confidence intervals in each time instant capturing uncertainty of the survey data (probabilistic estimation). Using the same set of obtained model parameters, we also provide a prediction over the next few years with 95% confidence intervals (probabilistic prediction). This technique is applied to a dynamic social model describing the evolution of the attitude of the Basque Country population towards the revolutionary organization ETA.
|
A probabilistic estimation and prediction technique for dynamic
continuous social science models: The evolution of the attitude of the Basque
Country population towards ETA as a case study
| 2,610
|
The true process that generated data cannot be determined when multiple explanations are possible. Prediction requires a model of the probability that a process, chosen randomly from the set of candidate explanations, generates some future observation. The best model includes all of the information contained in the minimal description of the data that is not contained in the data. It is closely related to the Halting Problem and is logarithmic in the size of the data. Prediction is difficult because the ideal model is not computable, and the best computable model is not "findable." However, the error from any approximation can be bounded by the size of the description using the model.
|
The Least Wrong Model Is Not in the Data
| 2,611
|
The Bayesian Classification represents a supervised learning method as well as a statistical method for classification. Assumes an underlying probabilistic model and it allows us to capture uncertainty about the model in a principled way by determining probabilities of the outcomes. This Classification is named after Thomas Bayes (1702-1761), who proposed the Bayes Theorem. Bayesian classification provides practical learning algorithms and prior knowledge and observed data can be combined. Bayesian Classification provides a useful perspective for understanding and evaluating many learning algorithms. It calculates explicit probabilities for hypothesis and it is robust to noise in input data. In statistical classification the Bayes classifier minimises the probability of misclassification. That was a visual intuition for a simple case of the Bayes classifier, also called: 1)Idiot Bayes 2)Naive Bayes 3)Simple Bayes
|
Bayes and Naive Bayes Classifier
| 2,612
|
In this paper, we evaluate the performance of various parallel optimization methods for Kernel Support Vector Machines on multicore CPUs and GPUs. In particular, we provide the first comparison of algorithms with explicit and implicit parallelization. Most existing parallel implementations for multi-core or GPU architectures are based on explicit parallelization of Sequential Minimal Optimization (SMO)---the programmers identified parallelizable components and hand-parallelized them, specifically tuned for a particular architecture. We compare these approaches with each other and with implicitly parallelized algorithms---where the algorithm is expressed such that most of the work is done within few iterations with large dense linear algebra operations. These can be computed with highly-optimized libraries, that are carefully parallelized for a large variety of parallel platforms. We highlight the advantages and disadvantages of both approaches and compare them on various benchmark data sets. We find an approximate implicitly parallel algorithm which is surprisingly efficient, permits a much simpler implementation, and leads to unprecedented speedups in SVM training.
|
Parallel Support Vector Machines in Practice
| 2,613
|
We present the \textit{hierarchical Dirichlet scaling process} (HDSP), a Bayesian nonparametric mixed membership model. The HDSP generalizes the hierarchical Dirichlet process (HDP) to model the correlation structure between metadata in the corpus and mixture components. We construct the HDSP based on the normalized gamma representation of the Dirichlet process, and this construction allows incorporating a scaling function that controls the membership probabilities of the mixture components. We develop two scaling methods to demonstrate that different modeling assumptions can be expressed in the HDSP. We also derive the corresponding approximate posterior inference algorithms using variational Bayes. Through experiments on datasets of newswire, medical journal articles, conference proceedings, and product reviews, we show that the HDSP results in a better predictive performance than labeled LDA, partially labeled LDA, and author topic model and a better negative review classification performance than the supervised topic model and SVM.
|
Hierarchical Dirichlet Scaling Process
| 2,614
|
In this paper we have focused on an efficient feature selection method in classification of audio files. The main objective is feature selection and extraction. We have selected a set of features for further analysis, which represents the elements in feature vector. By extraction method we can compute a numerical representation that can be used to characterize the audio using the existing toolbox. In this study Gain Ratio (GR) is used as a feature selection measure. GR is used to select splitting attribute which will separate the tuples into different classes. The pulse clarity is considered as a subjective measure and it is used to calculate the gain of features of audio files. The splitting criterion is employed in the application to identify the class or the music genre of a specific audio file from testing database. Experimental results indicate that by using GR the application can produce a satisfactory result for music genre classification. After dimensionality reduction best three features have been selected out of various features of audio file and in this technique we will get more than 90% successful classification result.
|
An Efficient Feature Selection in Classification of Audio Files
| 2,615
|
Analysis of high dimensional noisy data is of essence across a variety of research fields. Feature selection techniques are designed to find the relevant feature subset that can facilitate classification or pattern detection. Traditional (supervised) feature selection methods utilize label information to guide the identification of relevant feature subsets. In this paper, however, we consider the unsupervised feature selection problem. Without the label information, it is particularly difficult to identify a small set of relevant features due to the noisy nature of real-world data which corrupts the intrinsic structure of the data. Our Gradient-based Laplacian Feature Selection (GLFS) selects important features by minimizing the variance of the Laplacian regularized least squares regression model. With $\ell_1$ relaxation, GLFS can find a sparse subset of features that is relevant to the Laplacian manifolds. Extensive experiments on simulated, three real-world object recognition and two computational biology datasets, have illustrated the power and superior performance of our approach over multiple state-of-the-art unsupervised feature selection methods. Additionally, we show that GLFS selects a sparser set of more relevant features in a supervised setting outperforming the popular elastic net methodology.
|
Gradient-based Laplacian Feature Selection
| 2,616
|
A traditional and intuitively appealing Multi-Task Multiple Kernel Learning (MT-MKL) method is to optimize the sum (thus, the average) of objective functions with (partially) shared kernel function, which allows information sharing amongst tasks. We point out that the obtained solution corresponds to a single point on the Pareto Front (PF) of a Multi-Objective Optimization (MOO) problem, which considers the concurrent optimization of all task objectives involved in the Multi-Task Learning (MTL) problem. Motivated by this last observation and arguing that the former approach is heuristic, we propose a novel Support Vector Machine (SVM) MT-MKL framework, that considers an implicitly-defined set of conic combinations of task objectives. We show that solving our framework produces solutions along a path on the aforementioned PF and that it subsumes the optimization of the average of objective functions as a special case. Using algorithms we derived, we demonstrate through a series of experimental results that the framework is capable of achieving better classification performance, when compared to other similar MTL approaches.
|
Pareto-Path Multi-Task Multiple Kernel Learning
| 2,617
|
This work presents a sound probabilistic method for enforcing adherence of the marginal probabilities of a multi-label model to automatically discovered deterministic relationships among labels. In particular we focus on discovering two kinds of relationships among the labels. The first one concerns pairwise positive entailement: pairs of labels, where the presence of one implies the presence of the other in all instances of a dataset. The second concerns exclusion: sets of labels that do not coexist in the same instances of the dataset. These relationships are represented with a Bayesian network. Marginal probabilities are entered as soft evidence in the network and adjusted through probabilistic inference. Our approach offers robust improvements in mean average precision compared to the standard binary relavance approach across all 12 datasets involved in our experiments. The discovery process helps interesting implicit knowledge to emerge, which could be useful in itself.
|
Discovering and Exploiting Entailment Relationships in Multi-Label
Learning
| 2,618
|
Ensemble classifier refers to a group of individual classifiers that are cooperatively trained on data set in a supervised classification problem. In this paper we present a review of commonly used ensemble classifiers in the literature. Some ensemble classifiers are also developed targeting specific applications. We also present some application driven ensemble classifiers in this paper.
|
Ensemble Classifiers and Their Applications: A Review
| 2,619
|
Consider a Machine Learning Service Provider (MLSP) designed to rapidly create highly accurate learners for a never-ending stream of new tasks. The challenge is to produce task-specific learners that can be trained from few labeled samples, even if tasks are not uniquely identified, and the number of tasks and input dimensionality are large. In this paper, we argue that the MLSP should exploit knowledge from previous tasks to build a good representation of the environment it is in, and more precisely, that useful representations for such a service are ones that minimize generalization error for a new hypothesis trained on a new task. We formalize this intuition with a novel method that minimizes an empirical proxy of the intra-task small-sample generalization error. We present several empirical results showing state-of-the art performance on single-task transfer, multitask learning, and the full lifelong learning problem.
|
Representation as a Service
| 2,620
|
Stochastic variational inference makes it possible to approximate posterior distributions induced by large datasets quickly using stochastic optimization. The algorithm relies on the use of fully factorized variational distributions. However, this "mean-field" independence approximation limits the fidelity of the posterior approximation, and introduces local optima. We show how to relax the mean-field approximation to allow arbitrary dependencies between global parameters and local hidden variables, producing better parameter estimates by reducing bias, sensitivity to local optima, and sensitivity to hyperparameters.
|
Structured Stochastic Variational Inference
| 2,621
|
Dropout and other feature noising schemes have shown promising results in controlling over-fitting by artificially corrupting the training data. Though extensive theoretical and empirical studies have been performed for generalized linear models, little work has been done for support vector machines (SVMs), one of the most successful approaches for supervised learning. This paper presents dropout training for linear SVMs. To deal with the intractable expectation of the non-smooth hinge loss under corrupting distributions, we develop an iteratively re-weighted least square (IRLS) algorithm by exploring data augmentation techniques. Our algorithm iteratively minimizes the expectation of a re-weighted least square problem, where the re-weights have closed-form solutions. The similar ideas are applied to develop a new IRLS algorithm for the expected logistic loss under corrupting distributions. Our algorithms offer insights on the connection and difference between the hinge loss and logistic loss in dropout training. Empirical results on several real datasets demonstrate the effectiveness of dropout training on significantly boosting the classification accuracy of linear SVMs.
|
Dropout Training for Support Vector Machines
| 2,622
|
Machine learning algorithms have been applied to predict agent behaviors in real-world dynamic systems, such as advertiser behaviors in sponsored search and worker behaviors in crowdsourcing. The behavior data in these systems are generated by live agents: once the systems change due to the adoption of the prediction models learnt from the behavior data, agents will observe and respond to these changes by changing their own behaviors accordingly. As a result, the behavior data will evolve and will not be identically and independently distributed, posing great challenges to the theoretical analysis on the machine learning algorithms for behavior prediction. To tackle this challenge, in this paper, we propose to use Markov Chain in Random Environments (MCRE) to describe the behavior data, and perform generalization analysis of the machine learning algorithms on its basis. Since the one-step transition probability matrix of MCRE depends on both previous states and the random environment, conventional techniques for generalization analysis cannot be directly applied. To address this issue, we propose a novel technique that transforms the original MCRE into a higher-dimensional time-homogeneous Markov chain. The new Markov chain involves more variables but is more regular, and thus easier to deal with. We prove the convergence of the new Markov chain when time approaches infinity. Then we prove a generalization bound for the machine learning algorithms on the behavior data generated by the new Markov chain, which depends on both the Markovian parameters and the covering number of the function class compounded by the loss function for behavior prediction and the behavior prediction model. To the best of our knowledge, this is the first work that performs the generalization analysis on data generated by complex processes in real-world dynamic systems.
|
Agent Behavior Prediction and Its Generalization Analysis
| 2,623
|
Multi-target regression is concerned with the simultaneous prediction of multiple continuous target variables based on the same set of input variables. It arises in several interesting industrial and environmental application domains, such as ecological modelling and energy forecasting. This paper presents an ensemble method for multi-target regression that constructs new target variables via random linear combinations of existing targets. We discuss the connection of our approach with multi-label classification algorithms, in particular RA$k$EL, which originally inspired this work, and a family of recent multi-label classification algorithms that involve output coding. Experimental results on 12 multi-target datasets show that it performs significantly better than a strong baseline that learns a single model for each target using gradient boosting and compares favourably to multi-objective random forest approach, which is a state-of-the-art approach. The experiments further show that our approach improves more when stronger unconditional dependencies exist among the targets.
|
Multi-Target Regression via Random Linear Target Combinations
| 2,624
|
Coactive learning is an online problem solving setting where the solutions provided by a solver are interactively improved by a domain expert, which in turn drives learning. In this paper we extend the study of coactive learning to problems where obtaining a globally optimal or near-optimal solution may be intractable or where an expert can only be expected to make small, local improvements to a candidate solution. The goal of learning in this new setting is to minimize the cost as measured by the expert effort over time. We first establish theoretical bounds on the average cost of the existing coactive Perceptron algorithm. In addition, we consider new online algorithms that use cost-sensitive and Passive-Aggressive (PA) updates, showing similar or improved theoretical bounds. We provide an empirical evaluation of the learners in various domains, which show that the Perceptron based algorithms are quite effective and that unlike the case for online classification, the PA algorithms do not yield significant performance gains.
|
Coactive Learning for Locally Optimal Problem Solving
| 2,625
|
Multi-view learning leverages correlations between different sources of data to make predictions in one view based on observations in another view. A popular approach is to assume that, both, the correlations between the views and the view-specific covariances have a low-rank structure, leading to inter-battery factor analysis, a model closely related to canonical correlation analysis. We propose a convex relaxation of this model using structured norm regularization. Further, we extend the convex formulation to a robust version by adding an l1-penalized matrix to our estimator, similarly to convex robust PCA. We develop and compare scalable algorithms for several convex multi-view models. We show experimentally that the view-specific correlations are improving data imputation performances, as well as labeling accuracy in real-world multi-label prediction tasks.
|
Overlapping Trace Norms in Multi-View Learning
| 2,626
|
In this paper, we present a learning method for sequence labeling tasks in which each example sequence has multiple label sequences. Our method learns multiple models, one model for each label sequence. Each model computes the joint probability of all label sequences given the example sequence. Although each model considers all label sequences, its primary focus is only one label sequence, and therefore, each model becomes a task-specific model, for the task belonging to that primary label. Such multiple models are learned {\it simultaneously} by facilitating the learning transfer among models through {\it explicit parameter sharing}. We experiment the proposed method on two applications and show that our method significantly outperforms the state-of-the-art method.
|
Multitask Learning for Sequence Labeling Tasks
| 2,627
|
Using an optimization algorithm to solve a machine learning problem is one of mainstreams in the field of science. In this work, we demonstrate a comprehensive comparison of some state-of-the-art first-order optimization algorithms for convex optimization problems in machine learning. We concentrate on several smooth and non-smooth machine learning problems with a loss function plus a regularizer. The overall experimental results show the superiority of primal-dual algorithms in solving a machine learning problem from the perspectives of the ease to construct, running time and accuracy.
|
A Comparison of First-order Algorithms for Machine Learning
| 2,628
|
A new method to represent and approximate rotation matrices is introduced. The method represents approximations of a rotation matrix $Q$ with linearithmic complexity, i.e. with $\frac{1}{2}n\lg(n)$ rotations over pairs of coordinates, arranged in an FFT-like fashion. The approximation is "learned" using gradient descent. It allows to represent symmetric matrices $H$ as $QDQ^T$ where $D$ is a diagonal matrix. It can be used to approximate covariance matrix of Gaussian models in order to speed up inference, or to estimate and track the inverse Hessian of an objective function by relating changes in parameters to changes in gradient along the trajectory followed by the optimization procedure. Experiments were conducted to approximate synthetic matrices, covariance matrices of real data, and Hessian matrices of objective functions involved in machine learning problems.
|
Fast Approximation of Rotations and Hessians matrices
| 2,629
|
In this paper, we propose to study four meteorological and seasonal time series coupled with a multi-layer perceptron (MLP) modeling. We chose to combine two transfer functions for the nodes of the hidden layer, and to use a temporal indicator (time index as input) in order to take into account the seasonal aspect of the studied time series. The results of the prediction concern two years of measurements and the learning step, eight independent years. We show that this methodology can improve the accuracy of meteorological data estimation compared to a classical MLP modelling with a homogenous transfer function.
|
Meteorological time series forecasting based on MLP modelling using
heterogeneous transfer functions
| 2,630
|
Hidden Markov models (HMMs) are widely used statistical models for modeling sequential data. The parameter estimation for HMMs from time series data is an important learning problem. The predominant methods for parameter estimation are based on local search heuristics, most notably the expectation-maximization (EM) algorithm. These methods are prone to local optima and oftentimes suffer from high computational and sample complexity. Recent years saw the emergence of spectral methods for the parameter estimation of HMMs, based on a method of moments approach. Two spectral learning algorithms as proposed by Hsu, Kakade and Zhang 2012 (arXiv:0811.4413) and Anandkumar, Hsu and Kakade 2012 (arXiv:1203.0683) are assessed in this work. Using experiments with synthetic data, the algorithms are compared with each other. Furthermore, the spectral methods are compared to the Baum-Welch algorithm, a well-established method applying the EM algorithm to HMMs. The spectral algorithms are found to have a much more favorable computational and sample complexity. Even though the algorithms readily handle high dimensional observation spaces, instability issues are encountered in this regime. In view of learning from real-world experimental data, the representation of real-valued observations for the use in spectral methods is discussed, presenting possible methods to represent data for the use in the learning algorithms.
|
Implementing spectral methods for hidden Markov models with real-valued
emissions
| 2,631
|
We investigate how different learning restrictions reduce learning power and how the different restrictions relate to one another. We give a complete map for nine different restrictions both for the cases of complete information learning and set-driven learning. This completes the picture for these well-studied \emph{delayable} learning restrictions. A further insight is gained by different characterizations of \emph{conservative} learning in terms of variants of \emph{cautious} learning. Our analyses greatly benefit from general theorems we give, for example showing that learners with exclusively delayable restrictions can always be assumed total.
|
A Map of Update Constraints in Inductive Inference
| 2,632
|
In this paper, we study the problem of learning a monotone DNF with at most $s$ terms of size (number of variables in each term) at most $r$ ($s$ term $r$-MDNF) from membership queries. This problem is equivalent to the problem of learning a general hypergraph using hyperedge-detecting queries, a problem motivated by applications arising in chemical reactions and genome sequencing. We first present new lower bounds for this problem and then present deterministic and randomized adaptive algorithms with query complexities that are almost optimal. All the algorithms we present in this paper run in time linear in the query complexity and the number of variables $n$. In addition, all of the algorithms we present in this paper are asymptotically tight for fixed $r$ and/or $s$.
|
On Exact Learning Monotone DNF from Membership Queries
| 2,633
|
We present a new algorithm for domain adaptation improving upon a discrepancy minimization algorithm previously shown to outperform a number of algorithms for this task. Unlike many previous algorithms for domain adaptation, our algorithm does not consist of a fixed reweighting of the losses over the training sample. We show that our algorithm benefits from a solid theoretical foundation and more favorable learning bounds than discrepancy minimization. We present a detailed description of our algorithm and give several efficient solutions for solving its optimization problem. We also report the results of several experiments showing that it outperforms discrepancy minimization.
|
Adaptation Algorithm and Theory Based on Generalized Discrepancy
| 2,634
|
We consider the problem of proper learning a Boolean Halfspace with integer weights $\{0,1,\ldots,t\}$ from membership queries only. The best known algorithm for this problem is an adaptive algorithm that asks $n^{O(t^5)}$ membership queries where the best lower bound for the number of membership queries is $n^t$ [Learning Threshold Functions with Small Weights Using Membership Queries. COLT 1999] In this paper we close this gap and give an adaptive proper learning algorithm with two rounds that asks $n^{O(t)}$ membership queries. We also give a non-adaptive proper learning algorithm that asks $n^{O(t^3)}$ membership queries.
|
Learning Boolean Halfspaces with Small Weights from Membership Queries
| 2,635
|
The fundamental theorem of statistical learning states that for binary classification problems, any Empirical Risk Minimization (ERM) learning rule has close to optimal sample complexity. In this paper we seek for a generic optimal learner for multiclass prediction. We start by proving a surprising result: a generic optimal multiclass learner must be improper, namely, it must have the ability to output hypotheses which do not belong to the hypothesis class, even though it knows that all the labels are generated by some hypothesis from the class. In particular, no ERM learner is optimal. This brings back the fundmamental question of "how to learn"? We give a complete answer to this question by giving a new analysis of the one-inclusion multiclass learner of Rubinstein et al (2006) showing that its sample complexity is essentially optimal. Then, we turn to study the popular hypothesis class of generalized linear classifiers. We derive optimal learners that, unlike the one-inclusion algorithm, are computationally efficient. Furthermore, we show that the sample complexity of these learners is better than the sample complexity of the ERM rule, thus settling in negative an open question due to Collins (2005).
|
Optimal Learners for Multiclass Problems
| 2,636
|
We prove the existence of a canonical form for semi-deterministic transducers with incomparable sets of output strings. Based on this, we develop an algorithm which learns semi-deterministic transducers given access to translation queries. We also prove that there is no learning algorithm for semi-deterministic transducers that uses only domain knowledge.
|
A Canonical Semi-Deterministic Transducer
| 2,637
|
We consider a reinforcement learning setting introduced in (Maillard et al., NIPS 2011) where the learner does not have explicit access to the states of the underlying Markov decision process (MDP). Instead, she has access to several models that map histories of past interactions to states. Here we improve over known regret bounds in this setting, and more importantly generalize to the case where the models given to the learner do not contain a true model resulting in an MDP representation but only approximations of it. We also give improved error bounds for state aggregation.
|
Selecting Near-Optimal Approximate State Representations in
Reinforcement Learning
| 2,638
|
We study the convex relaxation of clustering and hamming embedding, focusing on the asymmetric case (co-clustering and asymmetric hamming embedding), understanding their relationship to LSH as studied by (Charikar 2002) and to the max-norm ball, and the differences between their symmetric and asymmetric versions.
|
Clustering, Hamming Embedding, Generalized LSH and the Max Norm
| 2,639
|
We present algorithms for reducing the Dueling Bandits problem to the conventional (stochastic) Multi-Armed Bandits problem. The Dueling Bandits problem is an online model of learning with ordinal feedback of the form "A is preferred to B" (as opposed to cardinal feedback like "A has value 2.5"), giving it wide applicability in learning from implicit user feedback and revealed and stated preferences. In contrast to existing algorithms for the Dueling Bandits problem, our reductions -- named $\Doubler$, $\MultiSbm$ and $\DoubleSbm$ -- provide a generic schema for translating the extensive body of known results about conventional Multi-Armed Bandit algorithms to the Dueling Bandits setting. For $\Doubler$ and $\MultiSbm$ we prove regret upper bounds in both finite and infinite settings, and conjecture about the performance of $\DoubleSbm$ which empirically outperforms the other two as well as previous algorithms in our experiments. In addition, we provide the first almost optimal regret bound in terms of second order terms, such as the differences between the values of the arms.
|
Reducing Dueling Bandits to Cardinal Bandits
| 2,640
|
The logistic loss function is often advocated in machine learning and statistics as a smooth and strictly convex surrogate for the 0-1 loss. In this paper we investigate the question of whether these smoothness and convexity properties make the logistic loss preferable to other widely considered options such as the hinge loss. We show that in contrast to known asymptotic bounds, as long as the number of prediction/optimization iterations is sub exponential, the logistic loss provides no improvement over a generic non-smooth loss function such as the hinge loss. In particular we show that the convergence rate of stochastic logistic optimization is bounded from below by a polynomial in the diameter of the decision set and the number of prediction iterations, and provide a matching tight upper bound. This resolves the COLT open problem of McMahan and Streeter (2012).
|
Logistic Regression: Tight Bounds for Stochastic and Online Optimization
| 2,641
|
Dyadic prediction methods operate on pairs of objects (dyads), aiming to infer labels for out-of-sample dyads. We consider the full and almost full cold start problem in dyadic prediction, a setting that occurs when both objects in an out-of-sample dyad have not been observed during training, or if one of them has been observed, but very few times. A popular approach for addressing this problem is to train a model that makes predictions based on a pairwise feature representation of the dyads, or, in case of kernel methods, based on a tensor product pairwise kernel. As an alternative to such a kernel approach, we introduce a novel two-step learning algorithm that borrows ideas from the fields of pairwise learning and spectral filtering. We show theoretically that the two-step method is very closely related to the tensor product kernel approach, and experimentally that it yields a slightly better predictive performance. Moreover, unlike existing tensor product kernel methods, the two-step method allows closed-form solutions for training and parameter selection via cross-validation estimates both in the full and almost full cold start settings, making the approach much more efficient and straightforward to implement.
|
A two-step learning approach for solving full and almost full cold start
problems in dyadic prediction
| 2,642
|
We study a new class of online learning problems where each of the online algorithm's actions is assigned an adversarial value, and the loss of the algorithm at each step is a known and deterministic function of the values assigned to its recent actions. This class includes problems where the algorithm's loss is the minimum over the recent adversarial values, the maximum over the recent values, or a linear combination of the recent values. We analyze the minimax regret of this class of problems when the algorithm receives bandit feedback, and prove that when the minimum or maximum functions are used, the minimax regret is $\tilde \Omega(T^{2/3})$ (so called hard online learning problems), and when a linear function is used, the minimax regret is $\tilde O(\sqrt{T})$ (so called easy learning problems). Previously, the only online learning problem that was known to be provably hard was the multi-armed bandit with switching costs.
|
Online Learning with Composite Loss Functions
| 2,643
|
This paper concerns the distributed training of nonlinear kernel machines on Map-Reduce. We show that a re-formulation of Nystr\"om approximation based solution which is solved using gradient based techniques is well suited for this, especially when it is necessary to work with a large number of basis points. The main advantages of this approach are: avoidance of computing the pseudo-inverse of the kernel sub-matrix corresponding to the basis points; simplicity and efficiency of the distributed part of the computations; and, friendliness to stage-wise addition of basis points. We implement the method using an AllReduce tree on Hadoop and demonstrate its value on a few large benchmark datasets.
|
A Distributed Algorithm for Training Nonlinear Kernel Machines
| 2,644
|
Distributed training of $l_1$ regularized classifiers has received great attention recently. Most existing methods approach this problem by taking steps obtained from approximating the objective by a quadratic approximation that is decoupled at the individual variable level. These methods are designed for multicore and MPI platforms where communication costs are low. They are inefficient on systems such as Hadoop running on a cluster of commodity machines where communication costs are substantial. In this paper we design a distributed algorithm for $l_1$ regularization that is much better suited for such systems than existing algorithms. A careful cost analysis is used to support these points and motivate our method. The main idea of our algorithm is to do block optimization of many variables on the actual objective function within each computing node; this increases the computational cost per step that is matched with the communication cost, and decreases the number of outer iterations, thus yielding a faster overall method. Distributed Gauss-Seidel and Gauss-Southwell greedy schemes are used for choosing variables to update in each step. We establish global convergence theory for our algorithm, including Q-linear rate of convergence. Experiments on two benchmark problems show our method to be much faster than existing methods.
|
A distributed block coordinate descent method for training $l_1$
regularized linear classifiers
| 2,645
|
We consider stochastic multi-armed bandit problems where the expected reward is a Lipschitz function of the arm, and where the set of arms is either discrete or continuous. For discrete Lipschitz bandits, we derive asymptotic problem specific lower bounds for the regret satisfied by any algorithm, and propose OSLB and CKL-UCB, two algorithms that efficiently exploit the Lipschitz structure of the problem. In fact, we prove that OSLB is asymptotically optimal, as its asymptotic regret matches the lower bound. The regret analysis of our algorithms relies on a new concentration inequality for weighted sums of KL divergences between the empirical distributions of rewards and their true distributions. For continuous Lipschitz bandits, we propose to first discretize the action space, and then apply OSLB or CKL-UCB, algorithms that provably exploit the structure efficiently. This approach is shown, through numerical experiments, to significantly outperform existing algorithms that directly deal with the continuous set of arms. Finally the results and algorithms are extended to contextual bandits with similarities.
|
Lipschitz Bandits: Regret Lower Bounds and Optimal Algorithms
| 2,646
|
We present a new optimization-theoretic approach to analyzing Follow-the-Leader style algorithms, particularly in the setting where perturbations are used as a tool for regularization. We show that adding a strongly convex penalty function to the decision rule and adding stochastic perturbations to data correspond to deterministic and stochastic smoothing operations, respectively. We establish an equivalence between "Follow the Regularized Leader" and "Follow the Perturbed Leader" up to the smoothness properties. This intuition leads to a new generic analysis framework that recovers and improves the previous known regret bounds of the class of algorithms commonly known as Follow the Perturbed Leader.
|
Online Linear Optimization via Smoothing
| 2,647
|
Random Forest (RF) is a powerful ensemble method for classification and regression tasks. It consists of decision trees set. Although, a single tree is well interpretable for human, the ensemble of trees is a black-box model. The popular technique to look inside the RF model is to visualize a RF proximity matrix obtained on data samples with Multidimensional Scaling (MDS) method. Herein, we present a novel method based on Self-Organising Maps (SOM) for revealing intrinsic relationships in data that lay inside the RF used for classification tasks. We propose an algorithm to learn the SOM with the proximity matrix obtained from the RF. The visualization of RF proximity matrix with MDS and SOM is compared. What is more, the SOM learned with the RF proximity matrix has better classification accuracy in comparison to SOM learned with Euclidean distance. Presented approach enables better understanding of the RF and additionally improves accuracy of the SOM.
|
Visualizing Random Forest with Self-Organising Map
| 2,648
|
In this paper, we set forth a new vision of reinforcement learning developed by us over the past few years, one that yields mathematically rigorous solutions to longstanding important questions that have remained unresolved: (i) how to design reliable, convergent, and robust reinforcement learning algorithms (ii) how to guarantee that reinforcement learning satisfies pre-specified "safety" guarantees, and remains in a stable region of the parameter space (iii) how to design "off-policy" temporal difference learning algorithms in a reliable and stable manner, and finally (iv) how to integrate the study of reinforcement learning into the rich theory of stochastic optimization. In this paper, we provide detailed answers to all these questions using the powerful framework of proximal operators. The key idea that emerges is the use of primal dual spaces connected through the use of a Legendre transform. This allows temporal difference updates to occur in dual spaces, allowing a variety of important technical advantages. The Legendre transform elegantly generalizes past algorithms for solving reinforcement learning problems, such as natural gradient methods, which we show relate closely to the previously unconnected framework of mirror descent methods. Equally importantly, proximal operator theory enables the systematic development of operator splitting methods that show how to safely and reliably decompose complex products of gradients that occur in recent variants of gradient-based temporal difference learning. This key technical innovation makes it possible to finally design "true" stochastic gradient methods for reinforcement learning. Finally, Legendre transforms enable a variety of other benefits, including modeling sparsity and domain geometry. Our work builds extensively on recent work on the convergence of saddle-point algorithms, and on the theory of monotone operators.
|
Proximal Reinforcement Learning: A New Theory of Sequential Decision
Making in Primal-Dual Spaces
| 2,649
|
BayesOpt is a library with state-of-the-art Bayesian optimization methods to solve nonlinear optimization, stochastic bandits or sequential experimental design problems. Bayesian optimization is sample efficient by building a posterior distribution to capture the evidence and prior knowledge for the target function. Built in standard C++, the library is extremely efficient while being portable and flexible. It includes a common interface for C, C++, Python, Matlab and Octave.
|
BayesOpt: A Bayesian Optimization Library for Nonlinear Optimization,
Experimental Design and Bandits
| 2,650
|
K-means algorithm is a very popular clustering algorithm which is famous for its simplicity. Distance measure plays a very important rule on the performance of this algorithm. We have different distance measure techniques available. But choosing a proper technique for distance calculation is totally dependent on the type of the data that we are going to cluster. In this paper an experimental study is done in Matlab to cluster the iris and wine data sets with different distance measures and thereby observing the variation of the performances shown.
|
Effect of Different Distance Measures on the Performance of K-Means
Algorithm: An Experimental Study in Matlab
| 2,651
|
A useful strategy to deal with complex classification scenarios is the "divide and conquer" approach. The mixture of experts (MOE) technique makes use of this strategy by joinly training a set of classifiers, or experts, that are specialized in different regions of the input space. A global model, or gate function, complements the experts by learning a function that weights their relevance in different parts of the input space. Local feature selection appears as an attractive alternative to improve the specialization of experts and gate function, particularly, for the case of high dimensional data. Our main intuition is that particular subsets of dimensions, or subspaces, are usually more appropriate to classify instances located in different regions of the input space. Accordingly, this work contributes with a regularized variant of MoE that incorporates an embedded process for local feature selection using $L1$ regularization, with a simultaneous expert selection. The experiments are still pending.
|
Simultaneous Feature and Expert Selection within Mixture of Experts
| 2,652
|
The performance of prediction models is often based on "abstract metrics" that estimate the model's ability to limit residual errors between the observed and predicted values. However, meaningful evaluation and selection of prediction models for end-user domains requires holistic and application-sensitive performance measures. Inspired by energy consumption prediction models used in the emerging "big data" domain of Smart Power Grids, we propose a suite of performance measures to rationally compare models along the dimensions of scale independence, reliability, volatility and cost. We include both application independent and dependent measures, the latter parameterized to allow customization by domain experts to fit their scenario. While our measures are generalizable to other domains, we offer an empirical analysis using real energy use data for three Smart Grid applications: planning, customer education and demand response, which are relevant for energy sustainability. Our results underscore the value of the proposed measures to offer a deeper insight into models' behavior and their impact on real applications, which benefit both data mining researchers and practitioners.
|
Holistic Measures for Evaluating Prediction Models in Smart Grids
| 2,653
|
Information divergence that measures the difference between two nonnegative matrices or tensors has found its use in a variety of machine learning problems. Examples are Nonnegative Matrix/Tensor Factorization, Stochastic Neighbor Embedding, topic models, and Bayesian network optimization. The success of such a learning task depends heavily on a suitable divergence. A large variety of divergences have been suggested and analyzed, but very few results are available for an objective choice of the optimal divergence for a given task. Here we present a framework that facilitates automatic selection of the best divergence among a given family, based on standard maximum likelihood estimation. We first propose an approximated Tweedie distribution for the beta-divergence family. Selecting the best beta then becomes a machine learning problem solved by maximum likelihood. Next, we reformulate alpha-divergence in terms of beta-divergence, which enables automatic selection of alpha by maximum likelihood with reuse of the learning principle for beta-divergence. Furthermore, we show the connections between gamma and beta-divergences as well as R\'enyi and alpha-divergences, such that our automatic selection framework is extended to non-separable divergences. Experiments on both synthetic and real-world data demonstrate that our method can quite accurately select information divergence across different learning problems and various divergence families.
|
Learning the Information Divergence
| 2,654
|
In this paper we explore how machine learning techniques can be applied to the discovery of efficient mathematical identities. We introduce an attribute grammar framework for representing symbolic expressions. Given a set of grammar rules we build trees that combine different rules, looking for branches which yield compositions that are analytically equivalent to a target expression, but of lower computational complexity. However, as the size of the trees grows exponentially with the complexity of the target expression, brute force search is impractical for all but the simplest of expressions. Consequently, we introduce two novel learning approaches that are able to learn from simpler expressions to guide the tree search. The first of these is a simple n-gram model, the other being a recursive neural-network. We show how these approaches enable us to derive complex identities, beyond reach of brute-force search, or human derivation.
|
Learning to Discover Efficient Mathematical Identities
| 2,655
|
We study the problem of multiclass classification with an extremely large number of classes (k), with the goal of obtaining train and test time complexity logarithmic in the number of classes. We develop top-down tree construction approaches for constructing logarithmic depth trees. On the theoretical front, we formulate a new objective function, which is optimized at each node of the tree and creates dynamic partitions of the data which are both pure (in terms of class labels) and balanced. We demonstrate that under favorable conditions, we can construct logarithmic depth trees that have leaves with low label entropy. However, the objective function at the nodes is challenging to optimize computationally. We address the empirical problem with a new online decision tree construction procedure. Experiments demonstrate that this online algorithm quickly achieves improvement in test error compared to more common logarithmic training time approaches, which makes it a plausible method in computationally constrained large-k applications.
|
Logarithmic Time Online Multiclass prediction
| 2,656
|
Many machine learning applications involve jointly predicting multiple mutually dependent output variables. Learning to search is a family of methods where the complex decision problem is cast into a sequence of decisions via a search space. Although these methods have shown promise both in theory and in practice, implementing them has been burdensomely awkward. In this paper, we show the search space can be defined by an arbitrary imperative program, turning learning to search into a credit assignment compiler. Altogether with the algorithmic improvements for the compiler, we radically reduce the complexity of programming and the running time. We demonstrate the feasibility of our approach on multiple joint prediction tasks. In all cases, we obtain accuracies as high as alternative approaches, at drastically reduced execution and programming time.
|
A Credit Assignment Compiler for Joint Prediction
| 2,657
|
We provide a general mechanism to design online learning algorithms based on a minimax analysis within a drifting-games framework. Different online learning settings (Hedge, multi-armed bandit problems and online convex optimization) are studied by converting into various kinds of drifting games. The original minimax analysis for drifting games is then used and generalized by applying a series of relaxations, starting from choosing a convex surrogate of the 0-1 loss function. With different choices of surrogates, we not only recover existing algorithms, but also propose new algorithms that are totally parameter-free and enjoy other useful properties. Moreover, our drifting-games framework naturally allows us to study high probability bounds without resorting to any concentration results, and also a generalized notion of regret that measures how good the algorithm is compared to all but the top small fraction of candidates. Finally, we translate our new Hedge algorithm into a new adaptive boosting algorithm that is computationally faster as shown in experiments, since it ignores a large number of examples on each round.
|
A Drifting-Games Analysis for Online Learning and Applications to
Boosting
| 2,658
|
Training deep directed graphical models with many hidden variables and performing inference remains a major challenge. Helmholtz machines and deep belief networks are such models, and the wake-sleep algorithm has been proposed to train them. The wake-sleep algorithm relies on training not just the directed generative model but also a conditional generative model (the inference network) that runs backward from visible to latent, estimating the posterior distribution of latent given visible. We propose a novel interpretation of the wake-sleep algorithm which suggests that better estimators of the gradient can be obtained by sampling latent variables multiple times from the inference network. This view is based on importance sampling as an estimator of the likelihood, with the approximate inference network as a proposal distribution. This interpretation is confirmed experimentally, showing that better likelihood can be achieved with this reweighted wake-sleep procedure. Based on this interpretation, we propose that a sigmoidal belief network is not sufficiently powerful for the layers of the inference network in order to recover a good estimator of the posterior distribution of latent variables. Our experiments show that using a more powerful layer model, such as NADE, yields substantially better generative models.
|
Reweighted Wake-Sleep
| 2,659
|
Because reinforcement learning suffers from a lack of scalability, online value (and Q-) function approximation has received increasing interest this last decade. This contribution introduces a novel approximation scheme, namely the Kalman Temporal Differences (KTD) framework, that exhibits the following features: sample-efficiency, non-linear approximation, non-stationarity handling and uncertainty management. A first KTD-based algorithm is provided for deterministic Markov Decision Processes (MDP) which produces biased estimates in the case of stochastic transitions. Than the eXtended KTD framework (XKTD), solving stochastic MDP, is described. Convergence is analyzed for special cases for both deterministic and stochastic transitions. Related algorithms are experimented on classical benchmarks. They compare favorably to the state of the art while exhibiting the announced features.
|
Kalman Temporal Differences
| 2,660
|
Restricted Boltzmann machines (RBM) and its variants have become hot research topics recently, and widely applied to many classification problems, such as character recognition and document categorization. Often, classification RBM ignores the interclass relationship or prior knowledge of sharing information among classes. In this paper, we are interested in RBM with the hierarchical prior over classes. We assume parameters for nearby nodes are correlated in the hierarchical tree, and further the parameters at each node of the tree be orthogonal to those at its ancestors. We propose a hierarchical correlated RBM for classification problem, which generalizes the classification RBM with sharing information among different classes. In order to reduce the redundancy between node parameters in the hierarchy, we also introduce orthogonal restrictions to our objective function. We test our method on challenge datasets, and show promising results compared to competitive baselines.
|
Restricted Boltzmann Machine for Classification with Hierarchical
Correlated Prior
| 2,661
|
We evaluate the following Machine Learning techniques for Green Energy (Wind, Solar) Prediction: Bayesian Inference, Neural Networks, Support Vector Machines, Clustering techniques (PCA). Our objective is to predict green energy using weather forecasts, predict deviations from forecast green energy, find correlation amongst different weather parameters and green energy availability, recover lost or missing energy (/ weather) data. We use historical weather data and weather forecasts for the same.
|
Evaluation of Machine Learning Techniques for Green Energy Prediction
| 2,662
|
We study a sequential resource allocation problem involving a fixed number of recurring jobs. At each time-step the manager should distribute available resources among the jobs in order to maximise the expected number of completed jobs. Allocating more resources to a given job increases the probability that it completes, but with a cut-off. Specifically, we assume a linear model where the probability increases linearly until it equals one, after which allocating additional resources is wasteful. We assume the difficulty of each job is unknown and present the first algorithm for this problem and prove upper and lower bounds on its regret. Despite its apparent simplicity, the problem has a rich structure: we show that an appropriate optimistic algorithm can improve its learning speed dramatically beyond the results one normally expects for similar problems as the problem becomes resource-laden.
|
Optimal Resource Allocation with Semi-Bandit Feedback
| 2,663
|
Spectral learning recently generated lots of excitement in machine learning, largely because it is the first known method to produce consistent estimates (under suitable conditions) for several latent variable models. In contrast, maximum likelihood estimates may get trapped in local optima due to the non-convex nature of the likelihood function of latent variable models. In this paper, we do an empirical evaluation of spectral learning (SL) and expectation maximization (EM), which reveals an important gap between the theory and the practice. First, SL often leads to negative probabilities. Second, EM often yields better estimates than spectral learning and it does not seem to get stuck in local optima. We discuss how the rank of the model parameters and the amount of training data can yield negative probabilities. We also question the common belief that maximum likelihood estimators are necessarily inconsistent.
|
A Sober Look at Spectral Learning
| 2,664
|
Data mining research into time series classification (TSC) has focussed on alternative distance measures for nearest neighbour classifiers. It is standard practice to use 1-NN with Euclidean or dynamic time warping (DTW) distance as a straw man for comparison. As part of a wider investigation into elastic distance measures for TSC~\cite{lines14elastic}, we perform a series of experiments to test whether this standard practice is valid. Specifically, we compare 1-NN classifiers with Euclidean and DTW distance to standard classifiers, examine whether the performance of 1-NN Euclidean approaches that of 1-NN DTW as the number of cases increases, assess whether there is any benefit of setting $k$ for $k$-NN through cross validation whether it is worth setting the warping path for DTW through cross validation and finally is it better to use a window or weighting for DTW. Based on experiments on 77 problems, we conclude that 1-NN with Euclidean distance is fairly easy to beat but 1-NN with DTW is not, if window size is set through cross validation.
|
An Experimental Evaluation of Nearest Neighbour Time Series
Classification
| 2,665
|
Dictionary learning is a branch of signal processing and machine learning that aims at finding a frame (called dictionary) in which some training data admits a sparse representation. The sparser the representation, the better the dictionary. The resulting dictionary is in general a dense matrix, and its manipulation can be computationally costly both at the learning stage and later in the usage of this dictionary, for tasks such as sparse coding. Dictionary learning is thus limited to relatively small-scale problems. In this paper, inspired by usual fast transforms, we consider a general dictionary structure that allows cheaper manipulation, and propose an algorithm to learn such dictionaries --and their fast implementation-- over training data. The approach is demonstrated experimentally with the factorization of the Hadamard matrix and with synthetic dictionary learning experiments.
|
Learning computationally efficient dictionaries and their implementation
as fast transforms
| 2,666
|
Symmetric positive semidefinite (SPSD) matrix approximation is an important problem with applications in kernel methods. However, existing SPSD matrix approximation methods such as the Nystr\"om method only have weak error bounds. In this paper we conduct in-depth studies of an SPSD matrix approximation model and establish strong relative-error bounds. We call it the prototype model for it has more efficient and effective extensions, and some of its extensions have high scalability. Though the prototype model itself is not suitable for large-scale data, it is still useful to study its properties, on which the analysis of its extensions relies. This paper offers novel theoretical analysis, efficient algorithms, and a highly accurate extension. First, we establish a lower error bound for the prototype model and improve the error bound of an existing column selection algorithm to match the lower bound. In this way, we obtain the first optimal column selection algorithm for the prototype model. We also prove that the prototype model is exact under certain conditions. Second, we develop a simple column selection algorithm with a provable error bound. Third, we propose a so-called spectral shifting model to make the approximation more accurate when the eigenvalues of the matrix decay slowly, and the improvement is theoretically quantified. The spectral shifting method can also be applied to improve other SPSD matrix approximation models.
|
SPSD Matrix Approximation vis Column Selection: Theories, Algorithms,
and Extensions
| 2,667
|
We study the bandit problem where arms are associated with stationary phi-mixing processes and where rewards are therefore dependent: the question that arises from this setting is that of recovering some independence by ignoring the value of some rewards. As we shall see, the bandit problem we tackle requires us to address the exploration/exploitation/independence trade-off. To do so, we provide a UCB strategy together with a general regret analysis for the case where the size of the independence blocks (the ignored rewards) is fixed and we go a step beyond by providing an algorithm that is able to compute the size of the independence blocks from the data. Finally, we give an analysis of our bandit problem in the restless case, i.e., in the situation where the time counters for all mixing processes simultaneously evolve.
|
Stationary Mixing Bandits
| 2,668
|
In this research we address the problem of capturing recurring concepts in a data stream environment. Recurrence capture enables the re-use of previously learned classifiers without the need for re-learning while providing for better accuracy during the concept recurrence interval. We capture concepts by applying the Discrete Fourier Transform (DFT) to Decision Tree classifiers to obtain highly compressed versions of the trees at concept drift points in the stream and store such trees in a repository for future use. Our empirical results on real world and synthetic data exhibiting varying degrees of recurrence show that the Fourier compressed trees are more robust to noise and are able to capture recurring concepts with higher precision than a meta learning approach that chooses to re-use classifiers in their originally occurring form.
|
Mining Recurrent Concepts in Data Streams using the Discrete Fourier
Transform
| 2,669
|
Mixability is a property of a loss which characterizes when fast convergence is possible in the game of prediction with expert advice. We show that a key property of mixability generalizes, and the exp and log operations present in the usual theory are not as special as one might have thought. In doing this we introduce a more general notion of $\Phi$-mixability where $\Phi$ is a general entropy (\ie, any convex function on probabilities). We show how a property shared by the convex dual of any such entropy yields a natural algorithm (the minimizer of a regret bound) which, analogous to the classical aggregating algorithm, is guaranteed a constant regret when used with $\Phi$-mixable losses. We characterize precisely which $\Phi$ have $\Phi$-mixable losses and put forward a number of conjectures about the optimality and relationships between different choices of entropy.
|
Generalized Mixability via Entropic Duality
| 2,670
|
Learning the parameters of graphical models using the maximum likelihood estimation is generally hard which requires an approximation. Maximum composite likelihood estimations are statistical approximations of the maximum likelihood estimation which are higher-order generalizations of the maximum pseudo-likelihood estimation. In this paper, we propose a composite likelihood method and investigate its property. Furthermore, we apply our composite likelihood method to restricted Boltzmann machines.
|
Composite Likelihood Estimation for Restricted Boltzmann machines
| 2,671
|
The explosion in the amount of data available for analysis often necessitates a transition from batch to incremental clustering methods, which process one element at a time and typically store only a small subset of the data. In this paper, we initiate the formal analysis of incremental clustering methods focusing on the types of cluster structure that they are able to detect. We find that the incremental setting is strictly weaker than the batch model, proving that a fundamental class of cluster structures that can readily be detected in the batch setting is impossible to identify using any incremental method. Furthermore, we show how the limitations of incremental clustering can be overcome by allowing additional clusters.
|
Incremental Clustering: The Case for Extra Clusters
| 2,672
|
This paper examines the efficacy of different optimization techniques in a primal formulation of a support vector machine (SVM). Three main techniques are compared. The dataset used to compare all three techniques was the Sentiment Analysis on Movie Reviews dataset, from kaggle.com.
|
Comparison of SVM Optimization Techniques in the Primal
| 2,673
|
Structure learning of Conditional Random Fields (CRFs) can be cast into an L1-regularized optimization problem. To avoid optimizing over a fully linked model, gain-based or gradient-based feature selection methods start from an empty model and incrementally add top ranked features to it. However, for high-dimensional problems like statistical relational learning, training time of these incremental methods can be dominated by the cost of evaluating the gain or gradient of a large collection of candidate features. In this study we propose a fast feature evaluation algorithm called Contrastive Feature Induction (CFI), which only evaluates a subset of features that involve both variables with high signals (deviation from mean) and variables with high errors (residue). We prove that the gradient of candidate features can be represented solely as a function of signals and errors, and that CFI is an efficient approximation of gradient-based evaluation methods. Experiments on synthetic and real data sets show competitive learning speed and accuracy of CFI on pairwise CRFs, compared to state-of-the-art structure learning methods such as full optimization over all features, and Grafting.
|
Contrastive Feature Induction for Efficient Structure Learning of
Conditional Random Fields
| 2,674
|
We consider stochastic bandit problems with a continuous set of arms and where the expected reward is a continuous and unimodal function of the arm. No further assumption is made regarding the smoothness and the structure of the expected reward function. For these problems, we propose the Stochastic Pentachotomy (SP) algorithm, and derive finite-time upper bounds on its regret and optimization error. In particular, we show that, for any expected reward function $\mu$ that behaves as $\mu(x)=\mu(x^\star)-C|x-x^\star|^\xi$ locally around its maximizer $x^\star$ for some $\xi, C>0$, the SP algorithm is order-optimal. Namely its regret and optimization error scale as $O(\sqrt{T\log(T)})$ and $O(\sqrt{\log(T)/T})$, respectively, when the time horizon $T$ grows large. These scalings are achieved without the knowledge of $\xi$ and $C$. Our algorithm is based on asymptotically optimal sequential statistical tests used to successively trim an interval that contains the best arm with high probability. To our knowledge, the SP algorithm constitutes the first sequential arm selection rule that achieves a regret and optimization error scaling as $O(\sqrt{T})$ and $O(1/\sqrt{T})$, respectively, up to a logarithmic factor for non-smooth expected reward functions, as well as for smooth functions with unknown smoothness.
|
Unimodal Bandits without Smoothness
| 2,675
|
Two types of low cost-per-iteration gradient descent methods have been extensively studied in parallel. One is online or stochastic gradient descent (OGD/SGD), and the other is randomzied coordinate descent (RBCD). In this paper, we combine the two types of methods together and propose online randomized block coordinate descent (ORBCD). At each iteration, ORBCD only computes the partial gradient of one block coordinate of one mini-batch samples. ORBCD is well suited for the composite minimization problem where one function is the average of the losses of a large number of samples and the other is a simple regularizer defined on high dimensional variables. We show that the iteration complexity of ORBCD has the same order as OGD or SGD. For strongly convex functions, by reducing the variance of stochastic gradients, we show that ORBCD can converge at a geometric rate in expectation, matching the convergence rate of SGD with variance reduction and RBCD.
|
Randomized Block Coordinate Descent for Online and Stochastic
Optimization
| 2,676
|
Which ads should we display in sponsored search in order to maximize our revenue? How should we dynamically rank information sources to maximize the value of the ranking? These applications exhibit strong diminishing returns: Redundancy decreases the marginal utility of each ad or information source. We show that these and other problems can be formalized as repeatedly selecting an assignment of items to positions to maximize a sequence of monotone submodular functions that arrive one by one. We present an efficient algorithm for this general problem and analyze it in the no-regret model. Our algorithm possesses strong theoretical guarantees, such as a performance ratio that converges to the optimal constant of 1 - 1/e. We empirically evaluate our algorithm on two real-world online optimization problems on the web: ad allocation with submodular utilities, and dynamically ranking blogs to detect information cascades. Finally, we present a second algorithm that handles the more general case in which the feasible sets are given by a matroid constraint, while still maintaining a 1 - 1/e asymptotic performance ratio.
|
Online Submodular Maximization under a Matroid Constraint with
Application to Learning Assignments
| 2,677
|
Multi-label learning deals with the classification problems where each instance can be assigned with multiple labels simultaneously. Conventional multi-label learning approaches mainly focus on exploiting label correlations. It is usually assumed, explicitly or implicitly, that the label sets for training instances are fully labeled without any missing labels. However, in many real-world multi-label datasets, the label assignments for training instances can be incomplete. Some ground-truth labels can be missed by the labeler from the label set. This problem is especially typical when the number instances is very large, and the labeling cost is very high, which makes it almost impossible to get a fully labeled training set. In this paper, we study the problem of large-scale multi-label learning with incomplete label assignments. We propose an approach, called MPU, based upon positive and unlabeled stochastic gradient descent and stacked models. Unlike prior works, our method can effectively and efficiently consider missing labels and label correlations simultaneously, and is very scalable, that has linear time complexities over the size of the data. Extensive experiments on two real-world multi-label datasets show that our MPU model consistently outperform other commonly-used baselines.
|
Large-Scale Multi-Label Learning with Incomplete Label Assignments
| 2,678
|
Many problems in real-world applications involve predicting several random variables which are statistically related. Markov random fields (MRFs) are a great mathematical tool to encode such relationships. The goal of this paper is to combine MRFs with deep learning algorithms to estimate complex representations while taking into account the dependencies between the output random variables. Towards this goal, we propose a training algorithm that is able to learn structured models jointly with deep features that form the MRF potentials. Our approach is efficient as it blends learning and inference and makes use of GPU acceleration. We demonstrate the effectiveness of our algorithm in the tasks of predicting words from noisy images, as well as multi-class classification of Flickr photographs. We show that joint learning of the deep features and the MRF parameters results in significant performance gains.
|
Learning Deep Structured Models
| 2,679
|
This document describes a novel learning algorithm that classifies "bags" of instances rather than individual instances. A bag is labeled positive if it contains at least one positive instance (which may or may not be specifically identified), and negative otherwise. This class of problems is known as multi-instance learning problems, and is useful in situations where the class label at an instance level may be unavailable or imprecise or difficult to obtain, or in situations where the problem is naturally posed as one of classifying instance groups. The algorithm described here is an ensemble-based method, wherein the members of the ensemble are lazy learning classifiers learnt using the Citation Nearest Neighbour method. Diversity among the ensemble members is achieved by optimizing their parameters using a multi-objective optimization method, with the objectives being to maximize Class 1 accuracy and minimize false positive rate. The method has been found to be effective on the Musk1 benchmark dataset.
|
A multi-instance learning algorithm based on a stacked ensemble of lazy
learners
| 2,680
|
The subspace Restricted Boltzmann Machine (subspaceRBM) is a third-order Boltzmann machine where multiplicative interactions are between one visible and two hidden units. There are two kinds of hidden units, namely, gate units and subspace units. The subspace units reflect variations of a pattern in data and the gate unit is responsible for activating the subspace units. Additionally, the gate unit can be seen as a pooling feature. We evaluate the behavior of subspaceRBM through experiments with MNIST digit recognition task, measuring reconstruction error and classification error.
|
Subspace Restricted Boltzmann Machine
| 2,681
|
No matter the expressive power and sophistication of supervised learning algorithms, their effectiveness is restricted by the features describing the data. This is not a new insight in ML and many methods for feature selection, transformation, and construction have been developed. But while this is on-going for general techniques for feature selection and transformation, i.e. dimensionality reduction, work on feature construction, i.e. enriching the data, is by now mainly the domain of image, particularly character, recognition, and NLP. In this work, we propose a new general framework for feature construction. The need for feature construction in a data set is indicated by class outliers and discriminative pattern mining used to derive features on their k-neighborhoods. We instantiate the framework with LOF and C4.5-Rules, and evaluate the usefulness of the derived features on a diverse collection of UCI data sets. The derived features are more often useful than ones derived by DC-Fringe, and our approach is much less likely to overfit. But while a weak learner, Naive Bayes, benefits strongly from the feature construction, the effect is less pronounced for C4.5, and almost vanishes for an SVM leaner. Keywords: feature construction, classification, outlier detection
|
A feature construction framework based on outlier detection and
discriminative pattern mining
| 2,682
|
In the last several years, the intimate connection between convex optimization and learning problems, in both statistical and sequential frameworks, has shifted the focus of algorithmic machine learning to examine this interplay. In particular, on one hand, this intertwinement brings forward new challenges in reassessment of the performance of learning algorithms including generalization and regret bounds under the assumptions imposed by convexity such as analytical properties of loss functions (e.g., Lipschitzness, strong convexity, and smoothness). On the other hand, emergence of datasets of an unprecedented size, demands the development of novel and more efficient optimization algorithms to tackle large-scale learning problems. The overarching goal of this thesis is to reassess the smoothness of loss functions in statistical learning, sequential prediction/online learning, and stochastic optimization and explicate its consequences. In particular we examine how smoothness of loss function could be beneficial or detrimental in these settings in terms of sample complexity, statistical consistency, regret analysis, and convergence rate, and investigate how smoothness can be leveraged to devise more efficient learning algorithms.
|
Exploiting Smoothness in Statistical Learning, Sequential Prediction,
and Stochastic Optimization
| 2,683
|
This paper presents a Fast Synchronization Clustering algorithm (FSynC), which is an improved version of SynC algorithm. In order to decrease the time complexity of the original SynC algorithm, we combine grid cell partitioning method and Red-Black tree to construct the near neighbor point set of every point. By simulated experiments of some artificial data sets and several real data sets, we observe that FSynC algorithm can often get less time than SynC algorithm for many kinds of data sets. At last, it gives some research expectations to popularize this algorithm.
|
A Fast Synchronization Clustering Algorithm
| 2,684
|
We consider sequential decision making in a setting where regret is measured with respect to a set of stateful reference policies, and feedback is limited to observing the rewards of the actions performed (the so called "bandit" setting). If either the reference policies are stateless rather than stateful, or the feedback includes the rewards of all actions (the so called "expert" setting), previous work shows that the optimal regret grows like $\Theta(\sqrt{T})$ in terms of the number of decision rounds $T$. The difficulty in our setting is that the decision maker unavoidably loses track of the internal states of the reference policies, and thus cannot reliably attribute rewards observed in a certain round to any of the reference policies. In fact, in this setting it is impossible for the algorithm to estimate which policy gives the highest (or even approximately highest) total reward. Nevertheless, we design an algorithm that achieves expected regret that is sublinear in $T$, of the form $O( T/\log^{1/4}{T})$. Our algorithm is based on a certain local repetition lemma that may be of independent interest. We also show that no algorithm can guarantee expected regret better than $O( T/\log^{3/2} T)$.
|
Chasing Ghosts: Competing with Stateful Policies
| 2,685
|
Many real-life data are described by categorical attributes without a pre-classification. A common data mining method used to extract information from this type of data is clustering. This method group together the samples from the data that are more similar than all other samples. But, categorical data pose a challenge when extracting information because: the calculation of two objects similarity is usually done by measuring the number of common features, but ignore a possible importance weighting; if the data may be divided differently according to different subsets of the features, the algorithm may find clusters with different meanings from each other, difficulting the post analysis. Data Co-Clustering of categorical data is the technique that tries to find subsets of samples that share a subset of features in common. By doing so, not only a sample may belong to more than one cluster but, the feature selection of each cluster describe its own characteristics. In this paper a novel Co-Clustering technique for categorical data is proposed by using Locality Sensitive Hashing technique in order to preprocess a list of Co-Clusters seeds based on a previous research. Results indicate this technique is capable of finding high quality Co-Clusters in many different categorical data sets and scales linearly with the data set size.
|
A Hash-based Co-Clustering Algorithm for Categorical Data
| 2,686
|
We propose to exploit {\em reconstruction} as a layer-local training signal for deep learning. Reconstructions can be propagated in a form of target propagation playing a role similar to back-propagation but helping to reduce the reliance on derivatives in order to perform credit assignment across many levels of possibly strong non-linearities (which is difficult for back-propagation). A regularized auto-encoder tends produce a reconstruction that is a more likely version of its input, i.e., a small move in the direction of higher likelihood. By generalizing gradients, target propagation may also allow to train deep networks with discrete hidden units. If the auto-encoder takes both a representation of input and target (or of any side information) in input, then its reconstruction of input representation provides a target towards a representation that is more likely, conditioned on all the side information. A deep auto-encoder decoding path generalizes gradient propagation in a learned way that can could thus handle not just infinitesimal changes but larger, discrete changes, hopefully allowing credit assignment through a long chain of non-linear operations. In addition to each layer being a good auto-encoder, the encoder also learns to please the upper layers by transforming the data into a space where it is easier to model by them, flattening manifolds and disentangling factors. The motivations and theoretical justifications for this approach are laid down in this paper, along with conjectures that will have to be verified either mathematically or experimentally, including a hypothesis stating that such auto-encoder mediated target propagation could play in brains the role of credit assignment through many non-linear, noisy and discrete transformations.
|
How Auto-Encoders Could Provide Credit Assignment in Deep Networks via
Target Propagation
| 2,687
|
With advances in data collection technologies, tensor data is assuming increasing prominence in many applications and the problem of supervised tensor learning has emerged as a topic of critical significance in the data mining and machine learning community. Conventional methods for supervised tensor learning mainly focus on learning kernels by flattening the tensor into vectors or matrices, however structural information within the tensors will be lost. In this paper, we introduce a new scheme to design structure-preserving kernels for supervised tensor learning. Specifically, we demonstrate how to leverage the naturally available structure within the tensorial representation to encode prior knowledge in the kernel. We proposed a tensor kernel that can preserve tensor structures based upon dual-tensorial mapping. The dual-tensorial mapping function can map each tensor instance in the input space to another tensor in the feature space while preserving the tensorial structure. Theoretically, our approach is an extension of the conventional kernels in the vector space to tensor space. We applied our novel kernel in conjunction with SVM to real-world tensor classification problems including brain fMRI classification for three different diseases (i.e., Alzheimer's disease, ADHD and brain damage by HIV). Extensive empirical studies demonstrate that our proposed approach can effectively boost tensor classification performances, particularly with small sample sizes.
|
DuSK: A Dual Structure-preserving Kernel for Supervised Tensor Learning
with Applications to Neuroimages
| 2,688
|
We define a general framework for a large class of combinatorial multi-armed bandit (CMAB) problems, where subsets of base arms with unknown distributions form super arms. In each round, a super arm is played and the base arms contained in the super arm are played and their outcomes are observed. We further consider the extension in which more based arms could be probabilistically triggered based on the outcomes of already triggered arms. The reward of the super arm depends on the outcomes of all played arms, and it only needs to satisfy two mild assumptions, which allow a large class of nonlinear reward instances. We assume the availability of an offline (\alpha,\beta)-approximation oracle that takes the means of the outcome distributions of arms and outputs a super arm that with probability {\beta} generates an {\alpha} fraction of the optimal expected reward. The objective of an online learning algorithm for CMAB is to minimize (\alpha,\beta)-approximation regret, which is the difference between the \alpha{\beta} fraction of the expected reward when always playing the optimal super arm, and the expected reward of playing super arms according to the algorithm. We provide CUCB algorithm that achieves O(log n) distribution-dependent regret, where n is the number of rounds played, and we further provide distribution-independent bounds for a large class of reward functions. Our regret analysis is tight in that it matches the bound of UCB1 algorithm (up to a constant factor) for the classical MAB problem, and it significantly improves the regret bound in a earlier paper on combinatorial bandits with linear rewards. We apply our CMAB framework to two new applications, probabilistic maximum coverage and social influence maximization, both having nonlinear reward structures. In particular, application to social influence maximization requires our extension on probabilistically triggered arms.
|
Combinatorial Multi-Armed Bandit and Its Extension to Probabilistically
Triggered Arms
| 2,689
|
Many of the ordinal regression models that have been proposed in the literature can be seen as methods that minimize a convex surrogate of the zero-one, absolute, or squared loss functions. A key property that allows to study the statistical implications of such approximations is that of Fisher consistency. Fisher consistency is a desirable property for surrogate loss functions and implies that in the population setting, i.e., if the probability distribution that generates the data were available, then optimization of the surrogate would yield the best possible model. In this paper we will characterize the Fisher consistency of a rich family of surrogate loss functions used in the context of ordinal regression, including support vector ordinal regression, ORBoosting and least absolute deviation. We will see that, for a family of surrogate loss functions that subsumes support vector ordinal regression and ORBoosting, consistency can be fully characterized by the derivative of a real-valued function at zero, as happens for convex margin-based surrogates in binary classification. We also derive excess risk bounds for a surrogate of the absolute error that generalize existing risk bounds for binary classification. Finally, our analysis suggests a novel surrogate of the squared error loss. We compare this novel surrogate with competing approaches on 9 different datasets. Our method shows to be highly competitive in practice, outperforming the least squares loss on 7 out of 9 datasets.
|
On the Consistency of Ordinal Regression Methods
| 2,690
|
We study the attainable regret for online linear optimization problems with bandit feedback, where unlike the full-information setting, the player can only observe its own loss rather than the full loss vector. We show that the price of bandit information in this setting can be as large as $d$, disproving the well-known conjecture that the regret for bandit linear optimization is at most $\sqrt{d}$ times the full-information regret. Surprisingly, this is shown using "trivial" modifications of standard domains, which have no effect in the full-information setting. This and other results we present highlight some interesting differences between full-information and bandit learning, which were not considered in previous literature.
|
On the Complexity of Bandit Linear Optimization
| 2,691
|
The VC dimension measures the capacity of a learning machine, and a low VC dimension leads to good generalization. While SVMs produce state-of-the-art learning performance, it is well known that the VC dimension of a SVM can be unbounded; despite good results in practice, there is no guarantee of good generalization. In this paper, we show how to learn a hyperplane classifier by minimizing an exact, or \boldmath{$\Theta$} bound on its VC dimension. The proposed approach, termed as the Minimal Complexity Machine (MCM), involves solving a simple linear programming problem. Experimental results show, that on a number of benchmark datasets, the proposed approach learns classifiers with error rates much less than conventional SVMs, while often using fewer support vectors. On many benchmark datasets, the number of support vectors is less than one-tenth the number used by SVMs, indicating that the MCM does indeed learn simpler representations.
|
Learning a hyperplane classifier by minimizing an exact bound on the VC
dimension
| 2,692
|
In this paper, a robust online sequential extreme learning machine (ROS-ELM) is proposed. It is based on the original OS-ELM with an adaptive selective ensemble framework. Two novel insights are proposed in this paper. First, a novel selective ensemble algorithm referred to as particle swarm optimization selective ensemble (PSOSEN) is proposed. Noting that PSOSEN is a general selective ensemble method which is applicable to any learning algorithms, including batch learning and online learning. Second, an adaptive selective ensemble framework for online learning is designed to balance the robustness and complexity of the algorithm. Experiments for both regression and classification problems with UCI data sets are carried out. Comparisons between OS-ELM, simple ensemble OS-ELM (EOS-ELM) and the proposed ROS-ELM empirically show that ROS-ELM significantly improves the robustness and stability.
|
Robust OS-ELM with a novel selective ensemble based on particle swarm
optimization
| 2,693
|
We propose a novel approach to sufficient dimension reduction in regression, based on estimating contour directions of negligible variation for the response surface. These directions span the orthogonal complement of the minimal space relevant for the regression, and can be extracted according to a measure of the variation in the response, leading to General Contour Regression(GCR). In comparison to exiisting sufficient dimension reduction techniques, this sontour-based mothology guarantees exhaustive estimation of the central space under ellipticity of the predictoor distribution and very mild additional assumptions, while maintaining vn-consisytency and somputational ease. Moreover, it proves to be robust to departures from ellipticity. We also establish some useful population properties for GCR. Simulations to compare performance with that of standard techniques such as ordinary least squares, sliced inverse regression, principal hessian directions, and sliced average variance estimation confirm the advntages anticipated by theoretical analyses. We also demonstrate the use of contour-based methods on a data set concerning grades of students from Massachusetts colleges.
|
Linear Contour Learning: A Method for Supervised Dimension Reduction
| 2,694
|
Vehicular sensor data consists of multiple time-series arising from a number of sensors. Using such multi-sensor data we would like to detect occurrences of specific events that vehicles encounter, e.g., corresponding to particular maneuvers that a vehicle makes or conditions that it encounters. Events are characterized by similar waveform patterns re-appearing within one or more sensors. Further such patterns can be of variable duration. In this work, we propose a method for detecting such events in time-series data using a novel feature descriptor motivated by similar ideas in image processing. We define the shape histogram: a constant dimension descriptor that nevertheless captures patterns of variable duration. We demonstrate the efficacy of using shape histograms as features to detect events in an SVM-based, multi-sensor, supervised learning scenario, i.e., multiple time-series are used to detect an event. We present results on real-life vehicular sensor data and show that our technique performs better than available pattern detection implementations on our data, and that it can also be used to combine features from multiple sensors resulting in better accuracy than using any single sensor. Since previous work on pattern detection in time-series has been in the single series context, we also present results using our technique on multiple standard time-series datasets and show that it is the most versatile in terms of how it ranks compared to other published results.
|
Multi-Sensor Event Detection using Shape Histograms
| 2,695
|
AFP Algorithm is a learning algorithm for Horn formulas. We show that it does not improve the complexity of AFP Algorithm, if after each negative counterexample more that just one refinements are performed. Moreover, a canonical normal form for Horn formulas is presented, and it is proved that the output formula of AFP Algorithm is in this normal form.
|
AFP Algorithm and a Canonical Normal Form for Horn Formulas
| 2,696
|
Traditionally, Multi-task Learning (MTL) models optimize the average of task-related objective functions, which is an intuitive approach and which we will be referring to as Average MTL. However, a more general framework, referred to as Conic MTL, can be formulated by considering conic combinations of the objective functions instead; in this framework, Average MTL arises as a special case, when all combination coefficients equal 1. Although the advantage of Conic MTL over Average MTL has been shown experimentally in previous works, no theoretical justification has been provided to date. In this paper, we derive a generalization bound for the Conic MTL method, and demonstrate that the tightest bound is not necessarily achieved, when all combination coefficients equal 1; hence, Average MTL may not always be the optimal choice, and it is important to consider Conic MTL. As a byproduct of the generalization bound, it also theoretically explains the good experimental results of previous relevant works. Finally, we propose a new Conic MTL model, whose conic combination coefficients minimize the generalization bound, instead of choosing them heuristically as has been done in previous methods. The rationale and advantage of our model is demonstrated and verified via a series of experiments by comparing with several other methods.
|
Conic Multi-Task Classification
| 2,697
|
We study the distributed computing setting in which there are multiple servers, each holding a set of points, who wish to compute functions on the union of their point sets. A key task in this setting is Principal Component Analysis (PCA), in which the servers would like to compute a low dimensional subspace capturing as much of the variance of the union of their point sets as possible. Given a procedure for approximate PCA, one can use it to approximately solve $\ell_2$-error fitting problems such as $k$-means clustering and subspace clustering. The essential properties of an approximate distributed PCA algorithm are its communication cost and computational efficiency for a given desired accuracy in downstream applications. We give new algorithms and analyses for distributed PCA which lead to improved communication and computational costs for $k$-means clustering and related problems. Our empirical study on real world data shows a speedup of orders of magnitude, preserving communication with only a negligible degradation in solution quality. Some of these techniques we develop, such as a general transformation from a constant success probability subspace embedding to a high success probability subspace embedding with a dimension and sparsity independent of the success probability, may be of independent interest.
|
Improved Distributed Principal Component Analysis
| 2,698
|
Although multi-label learning can deal with many problems with label ambiguity, it does not fit some real applications well where the overall distribution of the importance of the labels matters. This paper proposes a novel learning paradigm named \emph{label distribution learning} (LDL) for such kind of applications. The label distribution covers a certain number of labels, representing the degree to which each label describes the instance. LDL is a more general learning framework which includes both single-label and multi-label learning as its special cases. This paper proposes six working LDL algorithms in three ways: problem transformation, algorithm adaptation, and specialized algorithm design. In order to compare the performance of the LDL algorithms, six representative and diverse evaluation measures are selected via a clustering analysis, and the first batch of label distribution datasets are collected and made publicly available. Experimental results on one artificial and fifteen real-world datasets show clear advantages of the specialized algorithms, which indicates the importance of special design for the characteristics of the LDL problem.
|
Label Distribution Learning
| 2,699
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.