paper_id
string
title
string
paper_url
string
pdf_url
string
authors
list
abstract
large_string
track
string
primary_area
string
doi
string
volume
string
issue
string
pages
string
abstract_source
string
arxiv_id
string
arxiv_id_source
string
10.1609/aaai.v35i9.16977
Balanced Open Set Domain Adaptation via Centroid Alignment
https://ojs.aaai.org/index.php/AAAI/article/view/16977
https://ojs.aaai.org/index.php/AAAI/article/download/16977/16784
[ "Mengmeng Jing", "Jingjing Li", "Lei Zhu", "Zhengming Ding", "Ke Lu", "Yang Yang" ]
Open Set Domain Adaptation (OSDA) is a challenging domain adaptation setting which allows the existence of unknown classes on the target domain. Although existing OSDA methods are good at classifying samples of known classes, they ignore the classification ability for the unknown samples, making them unbalanced OSDA me...
main
Machine Learning
10.1609/aaai.v35i9.16977
35
9
8013-8020
official
null
null
10.1609/aaai.v35i9.16978
Linearly Replaceable Filters for Deep Network Channel Pruning
https://ojs.aaai.org/index.php/AAAI/article/view/16978
https://ojs.aaai.org/index.php/AAAI/article/download/16978/16785
[ "Donggyu Joo", "Eojindl Yi", "Sunghyun Baek", "Junmo Kim" ]
Convolutional neural networks (CNNs) have achieved remarkable results; however, despite the development of deep learning, practical user applications are fairly limited because heavy networks can be used solely with the latest hardware and software supports. Therefore, network pruning is gaining attention for general a...
main
Machine Learning
10.1609/aaai.v35i9.16978
35
9
8021-8029
official
null
null
10.1609/aaai.v35i9.16979
A Sample-Efficient Algorithm for Episodic Finite-Horizon MDP with Constraints
https://ojs.aaai.org/index.php/AAAI/article/view/16979
https://ojs.aaai.org/index.php/AAAI/article/download/16979/16786
[ "Krishna C. Kalagarla", "Rahul Jain", "Pierluigi Nuzzo" ]
Constrained Markov decision processes (CMDPs) formalize sequential decision-making problems whose objective is to minimize a cost function while satisfying constraints on various cost functions. In this paper, we consider the setting of episodic fixed-horizon CMDPs. We propose an online algorithm which leverages the li...
main
Machine Learning
10.1609/aaai.v35i9.16979
35
9
8030-8037
official
2009.11348
title_snapshot
10.1609/aaai.v35i9.16980
Winning Lottery Tickets in Deep Generative Models
https://ojs.aaai.org/index.php/AAAI/article/view/16980
https://ojs.aaai.org/index.php/AAAI/article/download/16980/16787
[ "Neha Mukund Kalibhat", "Yogesh Balaji", "Soheil Feizi" ]
The lottery ticket hypothesis suggests that sparse, sub-networks of a given neural network, if initialized properly, can be trained to reach comparable or even better performance to that of the original network. Prior works in lottery tickets have primarily focused on the supervised learning setup, with several papers ...
main
Machine Learning
10.1609/aaai.v35i9.16980
35
9
8038-8046
official
2010.02350
title_snapshot
10.1609/aaai.v35i9.16981
Exploration via State influence Modeling
https://ojs.aaai.org/index.php/AAAI/article/view/16981
https://ojs.aaai.org/index.php/AAAI/article/download/16981/16788
[ "Yongxin Kang", "Enmin Zhao", "Kai Li", "Junliang Xing" ]
This paper studies the challenging problem of reinforcement learning (RL) in hard exploration tasks with sparse rewards. It focuses on the exploration stage before the agent gets the first positive reward, in which case, traditional RL algorithms with simple exploration strategies often work poorly. Unlike previous met...
main
Machine Learning
10.1609/aaai.v35i9.16981
35
9
8047-8054
official
null
null
10.1609/aaai.v35i9.16982
Deep Probabilistic Canonical Correlation Analysis
https://ojs.aaai.org/index.php/AAAI/article/view/16982
https://ojs.aaai.org/index.php/AAAI/article/download/16982/16789
[ "Mahdi Karami", "Dale Schuurmans" ]
We propose a deep generative framework for multi-view learning based on a probabilistic interpretation of canonical correlation analysis (CCA). The model combines a linear multi-view layer in the latent space with deep generative networks as observation models, to decompose the variability in multiple views into a shar...
main
Machine Learning
10.1609/aaai.v35i9.16982
35
9
8055-8063
official
null
null
10.1609/aaai.v35i9.16943
Topology Distance: A Topology-Based Approach for Evaluating Generative Adversarial Networks
https://ojs.aaai.org/index.php/AAAI/article/view/16943
https://ojs.aaai.org/index.php/AAAI/article/download/16943/16750
[ "Danijela Horak", "Simiao Yu", "Gholamreza Salimi-Khorshidi" ]
Automatic evaluation of the goodness of Generative Adversarial Networks (GANs) has been a challenge for the field of machine learning. In this work, we propose a distance complementary to existing measures: Topology Distance (TD), the main idea behind which is to compare the geometric and topological features of the la...
main
Machine Learning
10.1609/aaai.v35i9.16943
35
9
7721-7728
official
2002.12054
title_snapshot
10.1609/aaai.v35i9.16944
Storage Fit Learning with Feature Evolvable Streams
https://ojs.aaai.org/index.php/AAAI/article/view/16944
https://ojs.aaai.org/index.php/AAAI/article/download/16944/16751
[ "Bo-Jian Hou", "Yu-Hu Yan", "Peng Zhao", "Zhi-Hua Zhou" ]
Feature evolvable learning has been widely studied in recent years where old features will vanish and new features will emerge when learning with streams. Conventional methods usually assume that a label will be revealed after prediction at each time step. However, in practice, this assumption may not hold whereas no l...
main
Machine Learning
10.1609/aaai.v35i9.16944
35
9
7729-7736
official
2007.11280
title_snapshot
10.1609/aaai.v35i9.16945
Reinforcement Learning Based Multi-Agent Resilient Control: From Deep Neural Networks to an Adaptive Law
https://ojs.aaai.org/index.php/AAAI/article/view/16945
https://ojs.aaai.org/index.php/AAAI/article/download/16945/16752
[ "Jian Hou", "Fangyuan Wang", "Lili Wang", "Zhiyong Chen" ]
Recent advances in Multi-agent Reinforcement Learning (MARL) have made it possible to implement various tasks in cooperative as well as competitive scenarios through trial and error, and deep neural networks. These successes motivate us to bring the mechanism of MARL into the Multi-agent Resilient Consensus (MARC) prob...
main
Machine Learning
10.1609/aaai.v35i9.16945
35
9
7737-7745
official
null
null
10.1609/aaai.v35i9.16946
Slimmable Generative Adversarial Networks
https://ojs.aaai.org/index.php/AAAI/article/view/16946
https://ojs.aaai.org/index.php/AAAI/article/download/16946/16753
[ "Liang Hou", "Zehuan Yuan", "Lei Huang", "Huawei Shen", "Xueqi Cheng", "Changhu Wang" ]
Generative adversarial networks (GANs) have achieved remarkable progress in recent years, but the continuously growing scale of models make them challenging to deploy widely in practical applications. In particular, for real-time generation tasks, different devices require generators of different sizes due to varying c...
main
Machine Learning
10.1609/aaai.v35i9.16946
35
9
7746-7753
official
2012.05660
title_snapshot
10.1609/aaai.v35i9.16947
Disentangled Representation Learning in Heterogeneous Information Network for Large-scale Android Malware Detection in the COVID-19 Era and Beyond
https://ojs.aaai.org/index.php/AAAI/article/view/16947
https://ojs.aaai.org/index.php/AAAI/article/download/16947/16754
[ "Shifu Hou", "Yujie Fan", "Mingxuan Ju", "Yanfang Ye", "Wenqiang Wan", "Kui Wang", "Yinming Mei", "Qi Xiong", "Fudong Shao" ]
In the fight against the COVID-19 pandemic, many social activities have moved online; society's overwhelming reliance on the complex cyberspace makes its security more important than ever. In this paper, we propose and develop an intelligent system named Dr.HIN to protect users against the evolving Android malware atta...
main
Machine Learning
10.1609/aaai.v35i9.16947
35
9
7754-7761
official
null
null
10.1609/aaai.v35i9.16948
Gaussian Process Priors for View-Aware Inference
https://ojs.aaai.org/index.php/AAAI/article/view/16948
https://ojs.aaai.org/index.php/AAAI/article/download/16948/16755
[ "Yuxin Hou", "Ari Heljakka", "Arno Solin" ]
While frame-independent predictions with deep neural networks have become the prominent solutions to many computer vision tasks, the potential benefits of utilizing correlations between frames have received less attention. Even though probabilistic machine learning provides the ability to encode correlation as prior kn...
main
Machine Learning
10.1609/aaai.v35i9.16948
35
9
7762-7770
official
1912.03249
title_snapshot
10.1609/aaai.v35i9.16949
Boosting Multi-task Learning Through Combination of Task Labels - with Applications in ECG Phenotyping
https://ojs.aaai.org/index.php/AAAI/article/view/16949
https://ojs.aaai.org/index.php/AAAI/article/download/16949/16756
[ "Ming-En Hsieh", "Vincent Tseng" ]
Multi-task learning has increased in importance due to its superior performance by learning multiple different tasks simultaneously and its ability to perform several different tasks using a single model. In medical phenotyping, task labels are costly to acquire and might contain a certain degree of label noise. This d...
main
Machine Learning
10.1609/aaai.v35i9.16949
35
9
7771-7779
official
null
null
10.1609/aaai.v35i9.16950
OPQ: Compressing Deep Neural Networks with One-shot Pruning-Quantization
https://ojs.aaai.org/index.php/AAAI/article/view/16950
https://ojs.aaai.org/index.php/AAAI/article/download/16950/16757
[ "Peng Hu", "Xi Peng", "Hongyuan Zhu", "Mohamed M. Sabry Aly", "Jie Lin" ]
As Deep Neural Networks (DNNs) usually are overparameterized and have millions of weight parameters, it is challenging to deploy these large DNN models on resource-constrained hardware platforms, e.g., smartphones. Numerous network compression methods such as pruning and quantization are proposed to reduce the model si...
main
Machine Learning
10.1609/aaai.v35i9.16950
35
9
7780-7788
official
2205.11141
title_snapshot
10.1609/aaai.v35i9.16951
Multi-scale Graph Fusion for Co-saliency Detection
https://ojs.aaai.org/index.php/AAAI/article/view/16951
https://ojs.aaai.org/index.php/AAAI/article/download/16951/16758
[ "Rongyao Hu", "Zhenyun Deng", "Xiaofeng Zhu" ]
The key challenge of co-saliency detection is to extract discriminative features to distinguish the common salient foregrounds from backgrounds in a group of relevant images. In this paper, we propose a new co-saliency detection framework which includes two strategies to improve the discriminative ability of the featur...
main
Machine Learning
10.1609/aaai.v35i9.16951
35
9
7789-7796
official
null
null
10.1609/aaai.v35i9.16952
Continual Learning by Using Information of Each Class Holistically
https://ojs.aaai.org/index.php/AAAI/article/view/16952
https://ojs.aaai.org/index.php/AAAI/article/download/16952/16759
[ "Wenpeng Hu", "Qi Qin", "Mengyu Wang", "Jinwen Ma", "Bing Liu" ]
Continual learning (CL) incrementally learns a sequence of tasks while solving the catastrophic forgetting (CF) problem. Existing methods mainly try to deal with CF directly. In this paper, we propose to avoid CF by considering the features of each class holistically rather than only the discriminative information for ...
main
Machine Learning
10.1609/aaai.v35i9.16952
35
9
7797-7805
official
null
null
10.1609/aaai.v35i9.16953
Predictive Adversarial Learning from Positive and Unlabeled Data
https://ojs.aaai.org/index.php/AAAI/article/view/16953
https://ojs.aaai.org/index.php/AAAI/article/download/16953/16760
[ "Wenpeng Hu", "Ran Le", "Bing Liu", "Feng Ji", "Jinwen Ma", "Dongyan Zhao", "Rui Yan" ]
This paper studies learning from positive and unlabeled examples, known as PU learning. It proposes a novel PU learning method called Predictive Adversarial Networks (PAN) based on GAN (Generative Adversarial Networks). GAN learns a generator to generate data (e.g., images) to fool a discriminator which tries to determ...
main
Machine Learning
10.1609/aaai.v35i9.16953
35
9
7806-7814
official
null
null
10.1609/aaai.v35i9.16954
Multidimensional Uncertainty-Aware Evidential Neural Networks
https://ojs.aaai.org/index.php/AAAI/article/view/16954
https://ojs.aaai.org/index.php/AAAI/article/download/16954/16761
[ "Yibo Hu", "Yuzhe Ou", "Xujiang Zhao", "Jin-Hee Cho", "Feng Chen" ]
Traditional deep neural networks (NNs) have significantly contributed to the state-of-the-art performance in the task of classification under various application domains. However, NNs have not considered inherent uncertainty in data associated with the class probabilities where misclassification under uncertainty may e...
main
Machine Learning
10.1609/aaai.v35i9.16954
35
9
7815-7822
official
2012.13676
title_snapshot
10.1609/aaai.v35i9.16955
Adversarial Defence by Diversified Simultaneous Training of Deep Ensembles
https://ojs.aaai.org/index.php/AAAI/article/view/16955
https://ojs.aaai.org/index.php/AAAI/article/download/16955/16762
[ "Bo Huang", "Zhiwei Ke", "Yi Wang", "Wei Wang", "Linlin Shen", "Feng Liu" ]
Learning-based classifiers are susceptible to adversarial examples. Existing defence methods are mostly devised on individual classifiers. Recent studies showed that it is viable to increase adversarial robustness by promoting diversity over an ensemble of models. In this paper, we propose adversarial defence by encour...
main
Machine Learning
10.1609/aaai.v35i9.16955
35
9
7823-7831
official
null
null
10.1609/aaai.v35i9.16956
Accelerating Continuous Normalizing Flow with Trajectory Polynomial Regularization
https://ojs.aaai.org/index.php/AAAI/article/view/16956
https://ojs.aaai.org/index.php/AAAI/article/download/16956/16763
[ "Han-Hsien Huang", "Mi-Yen Yeh" ]
In this paper, we propose an approach to effectively accelerating the computation of continuous normalizing flow (CNF), which has been proven to be a powerful tool for the tasks such as variational inference and density estimation. The training time cost of CNF can be extremely high because the required number of funct...
main
Machine Learning
10.1609/aaai.v35i9.16956
35
9
7832-7839
official
2012.04228
title_snapshot
10.1609/aaai.v35i9.16957
Attributes-Guided and Pure-Visual Attention Alignment for Few-Shot Recognition
https://ojs.aaai.org/index.php/AAAI/article/view/16957
https://ojs.aaai.org/index.php/AAAI/article/download/16957/16764
[ "Siteng Huang", "Min Zhang", "Yachen Kang", "Donglin Wang" ]
The purpose of few-shot recognition is to recognize novel categories with a limited number of labeled examples in each class. To encourage learning from a supplementary view, recent approaches have introduced auxiliary semantic modalities into effective metric-learning frameworks that aim to learn a feature similarity ...
main
Machine Learning
10.1609/aaai.v35i9.16957
35
9
7840-7847
official
2009.04724
title_snapshot
10.1609/aaai.v35i9.16958
Learning to Reweight Imaginary Transitions for Model-Based Reinforcement Learning
https://ojs.aaai.org/index.php/AAAI/article/view/16958
https://ojs.aaai.org/index.php/AAAI/article/download/16958/16765
[ "Wenzhen Huang", "Qiyue Yin", "Junge Zhang", "Kaiqi Huang" ]
Model-based reinforcement learning (RL) is more sample efficient than model-free RL by using imaginary trajectories generated by the learned dynamics model. When the model is inaccurate or biased, imaginary trajectories may be deleterious for training the action-value and policy functions. To alleviate such problem, th...
main
Machine Learning
10.1609/aaai.v35i9.16958
35
9
7848-7856
official
2104.04174
title_snapshot
10.1609/aaai.v35i9.16959
ACMo: Angle-Calibrated Moment Methods for Stochastic Optimization
https://ojs.aaai.org/index.php/AAAI/article/view/16959
https://ojs.aaai.org/index.php/AAAI/article/download/16959/16766
[ "Xunpeng Huang", "Runxin Xu", "Hao Zhou", "Zhe Wang", "Zhengyang Liu", "Lei Li" ]
Stochastic gradient descent (SGD) is a widely used method for its outstanding generalization ability and simplicity. Adaptive gradient methods have been proposed to further accelerate the optimization process. In this paper, we revisit existing adaptive gradient optimization methods with a new interpretation. Such new ...
main
Machine Learning
10.1609/aaai.v35i9.16959
35
9
7857-7864
official
2006.07065
title_snapshot
10.1609/aaai.v35i9.16960
Personalized Cross-Silo Federated Learning on Non-IID Data
https://ojs.aaai.org/index.php/AAAI/article/view/16960
https://ojs.aaai.org/index.php/AAAI/article/download/16960/16767
[ "Yutao Huang", "Lingyang Chu", "Zirui Zhou", "Lanjun Wang", "Jiangchuan Liu", "Jian Pei", "Yong Zhang" ]
Non-IID data present a tough challenge for federated learning. In this paper, we explore a novel idea of facilitating pairwise collaborations between clients with similar data. We propose FedAMP, a new method employing federated attentive message passing to facilitate similar clients to collaborate more. We establish t...
main
Machine Learning
10.1609/aaai.v35i9.16960
35
9
7865-7873
official
2007.03797
title_snapshot
10.1609/aaai.v35i9.16961
Reward-Biased Maximum Likelihood Estimation for Linear Stochastic Bandits
https://ojs.aaai.org/index.php/AAAI/article/view/16961
https://ojs.aaai.org/index.php/AAAI/article/download/16961/16768
[ "Yu-Heng Hung", "Ping-Chun Hsieh", "Xi Liu", "P. R. Kumar" ]
Modifying the reward-biased maximum likelihood method originally proposed in the adaptive control literature, we propose novel learning algorithms to handle the explore-exploit trade-off in linear bandits problems as well as generalized linear bandits problems. We develop novel index policies that we prove achieve orde...
main
Machine Learning
10.1609/aaai.v35i9.16961
35
9
7874-7882
official
2010.04091
title_snapshot
10.1609/aaai.v35i9.16962
Large Batch Optimization for Deep Learning Using New Complete Layer-Wise Adaptive Rate Scaling
https://ojs.aaai.org/index.php/AAAI/article/view/16962
https://ojs.aaai.org/index.php/AAAI/article/download/16962/16769
[ "Zhouyuan Huo", "Bin Gu", "Heng Huang" ]
Training deep neural networks using a large batch size has shown promising results and benefits many real-world applications. Warmup is one of nontrivial techniques to stabilize the convergence of large batch training. However, warmup is an empirical method and it is still unknown whether there is a better algorithm wi...
main
Machine Learning
10.1609/aaai.v35i9.16962
35
9
7883-7890
official
null
null
10.1609/aaai.v35i9.16923
Increasing Iterate Averaging for Solving Saddle-Point Problems
https://ojs.aaai.org/index.php/AAAI/article/view/16923
https://ojs.aaai.org/index.php/AAAI/article/download/16923/16730
[ "Yuan Gao", "Christian Kroer", "Donald Goldfarb" ]
Many problems in machine learning and game theory can be formulated as saddle-point problems, for which various first-order methods have been developed and proven efficient in practice. Under the general convex-concave assumption, most first-order methods only guarantee an ergodic convergence rate, that is, the uniform...
main
Machine Learning
10.1609/aaai.v35i9.16923
35
9
7537-7544
official
1903.10646
title_snapshot
10.1609/aaai.v35i9.16924
Uncertainty-Aware Multi-View Representation Learning
https://ojs.aaai.org/index.php/AAAI/article/view/16924
https://ojs.aaai.org/index.php/AAAI/article/download/16924/16731
[ "Yu Geng", "Zongbo Han", "Changqing Zhang", "Qinghua Hu" ]
Learning from different data views by exploring the underlying complementary information among them can endow the representation with stronger expressive ability. However, high-dimensional features tend to contain noise, and furthermore, quality of data usually varies for different samples (even for different views), i...
main
Machine Learning
10.1609/aaai.v35i9.16924
35
9
7545-7553
official
2201.05776
title_snapshot
10.1609/aaai.v35i9.16925
Justicia: A Stochastic SAT Approach to Formally Verify Fairness
https://ojs.aaai.org/index.php/AAAI/article/view/16925
https://ojs.aaai.org/index.php/AAAI/article/download/16925/16732
[ "Bishwamittra Ghosh", "Debabrota Basu", "Kuldeep S. Meel" ]
As a technology ML is oblivious to societal good or bad, and thus, the field of fair machine learning has stepped up to propose multiple mathematical definitions, algorithms, and systems to ensure different notions of fairness in ML applications. Given the multitude of propositions, it has become imperative to formally...
main
Machine Learning
10.1609/aaai.v35i9.16925
35
9
7554-7563
official
2009.06516
title_snapshot
10.1609/aaai.v35i9.16926
The Importance of Modeling Data Missingness in Algorithmic Fairness: A Causal Perspective
https://ojs.aaai.org/index.php/AAAI/article/view/16926
https://ojs.aaai.org/index.php/AAAI/article/download/16926/16733
[ "Naman Goel", "Alfonso Amayuelas", "Amit Deshpande", "Amit Sharma" ]
Training datasets for machine learning often have some form of missingness. For example, to learn a model for deciding whom to give a loan, the available training data includes individuals who were given a loan in the past, but not those who were not. This missingness, if ignored, nullifies any fairness guarantee of th...
main
Machine Learning
10.1609/aaai.v35i9.16926
35
9
7564-7573
official
2012.11448
title_snapshot
10.1609/aaai.v35i9.16927
Attribute-Guided Adversarial Training for Robustness to Natural Perturbations
https://ojs.aaai.org/index.php/AAAI/article/view/16927
https://ojs.aaai.org/index.php/AAAI/article/download/16927/16734
[ "Tejas Gokhale", "Rushil Anirudh", "Bhavya Kailkhura", "Jayaraman J. Thiagarajan", "Chitta Baral", "Yezhou Yang" ]
While existing work in robust deep learning has focused on small pixel-level norm-based perturbations, this may not account for perturbations encountered in several real world settings. In many such cases although test data might not be available, broad specifications about the types of perturbations (such as an unknow...
main
Machine Learning
10.1609/aaai.v35i9.16927
35
9
7574-7582
official
2012.01806
title_snapshot
10.1609/aaai.v35i9.16928
Efficient On-Chip Learning for Optical Neural Networks Through Power-Aware Sparse Zeroth-Order Optimization
https://ojs.aaai.org/index.php/AAAI/article/view/16928
https://ojs.aaai.org/index.php/AAAI/article/download/16928/16735
[ "Jiaqi Gu", "Chenghao Feng", "Zheng Zhao", "Zhoufeng Ying", "Ray T. Chen", "David Z. Pan" ]
Optical neural networks (ONNs) have demonstrated record-breaking potential in high-performance neuromorphic computing due to their ultra-high execution speed and low energy consumption. However, current learning protocols fail to provide scalable and efficient solutions to photonic circuit optimization in practical app...
main
Machine Learning
10.1609/aaai.v35i9.16928
35
9
7583-7591
official
2012.11148
title_snapshot
10.1609/aaai.v35i9.16929
Attentive Neural Point Processes for Event Forecasting
https://ojs.aaai.org/index.php/AAAI/article/view/16929
https://ojs.aaai.org/index.php/AAAI/article/download/16929/16736
[ "Yulong Gu" ]
Event sequence, where each event is associated with a marker and a timestamp, is increasingly ubiquitous in various applications. Accordingly, event forecasting emerges to be a crucial problem, which aims to predict the next event based on the historical sequence. In this paper, we propose ANPP, an Attentive Neural Poi...
main
Machine Learning
10.1609/aaai.v35i9.16929
35
9
7592-7600
official
null
null
10.1609/aaai.v35i9.16930
Revisiting Iterative Back-Translation from the Perspective of Compositional Generalization
https://ojs.aaai.org/index.php/AAAI/article/view/16930
https://ojs.aaai.org/index.php/AAAI/article/download/16930/16737
[ "Yinuo Guo", "Hualei Zhu", "Zeqi Lin", "Bei Chen", "Jian-Guang Lou", "Dongmei Zhang" ]
Human intelligence exhibits compositional generalization (i.e., the capacity to understand and produce unseen combinations of seen components), but current neural seq2seq models lack such ability. In this paper, we revisit iterative back-translation, a simple yet effective semi-supervised method, to investigate whether...
main
Machine Learning
10.1609/aaai.v35i9.16930
35
9
7601-7609
official
2012.04276
title_snapshot
10.1609/aaai.v35i9.16931
Controllable Guarantees for Fair Outcomes via Contrastive Information Estimation
https://ojs.aaai.org/index.php/AAAI/article/view/16931
https://ojs.aaai.org/index.php/AAAI/article/download/16931/16738
[ "Umang Gupta", "Aaron M Ferber", "Bistra Dilkina", "Greg Ver Steeg" ]
Controlling bias in training datasets is vital for ensuring equal treatment, or parity, between different groups in downstream applications. A naive solution is to transform the data so that it is statistically independent of group membership, but this may throw away too much information when a reasonable compromise be...
main
Machine Learning
10.1609/aaai.v35i9.16931
35
9
7610-7619
official
2101.04108
title_snapshot
10.1609/aaai.v35i9.16932
Towards Reusable Network Components by Learning Compatible Representations
https://ojs.aaai.org/index.php/AAAI/article/view/16932
https://ojs.aaai.org/index.php/AAAI/article/download/16932/16739
[ "Michael Gygli", "Jasper Uijlings", "Vittorio Ferrari" ]
This paper proposes to make a first step towards compatible and hence reusable network components. Rather than training networks for different tasks independently, we adapt the training process to produce network components that are compatible across tasks. In particular, we split a network into two components, a featu...
main
Machine Learning
10.1609/aaai.v35i9.16932
35
9
7620-7629
official
2004.03898
title_snapshot
10.1609/aaai.v35i9.16933
High-Dimensional Bayesian Optimization via Tree-Structured Additive Models
https://ojs.aaai.org/index.php/AAAI/article/view/16933
https://ojs.aaai.org/index.php/AAAI/article/download/16933/16740
[ "Eric Han", "Ishank Arora", "Jonathan Scarlett" ]
Bayesian Optimization (BO) has shown significant success in tackling expensive low-dimensional black-box optimization problems. Many optimization problems of interest are high-dimensional, and scaling BO to such settings remains an important challenge. In this paper, we consider generalized additive models in which low...
main
Machine Learning
10.1609/aaai.v35i9.16933
35
9
7630-7638
official
2012.13088
title_snapshot
10.1609/aaai.v35i9.16934
Explanation Consistency Training: Facilitating Consistency-Based Semi-Supervised Learning with Interpretability
https://ojs.aaai.org/index.php/AAAI/article/view/16934
https://ojs.aaai.org/index.php/AAAI/article/download/16934/16741
[ "Tao Han", "Wei-Wei Tu", "Yu-Feng Li" ]
Unlabeled data exploitation and interpretability are usually both required in reality. They, however, are conducted independently, and very few works try to connect the two. For unlabeled data exploitation, state-of-the-art semi-supervised learning (SSL) results have been achieved via encouraging the consistency of mod...
main
Machine Learning
10.1609/aaai.v35i9.16934
35
9
7639-7646
official
null
null
10.1609/aaai.v35i9.16935
DeepSynth: Automata Synthesis for Automatic Task Segmentation in Deep Reinforcement Learning
https://ojs.aaai.org/index.php/AAAI/article/view/16935
https://ojs.aaai.org/index.php/AAAI/article/download/16935/16742
[ "Mohammadhosein Hasanbeig", "Natasha Yogananda Jeppu", "Alessandro Abate", "Tom Melham", "Daniel Kroening" ]
This paper proposes DeepSynth, a method for effective training of deep Reinforcement Learning (RL) agents when the reward is sparse and non-Markovian, but at the same time progress towards the reward requires achieving an unknown sequence of high-level objectives. Our method employs a novel algorithm for synthesis of c...
main
Machine Learning
10.1609/aaai.v35i9.16935
35
9
7647-7656
official
1911.10244
title_snapshot
10.1609/aaai.v35i9.16936
Liquid Time-constant Networks
https://ojs.aaai.org/index.php/AAAI/article/view/16936
https://ojs.aaai.org/index.php/AAAI/article/download/16936/16743
[ "Ramin Hasani", "Mathias Lechner", "Alexander Amini", "Daniela Rus", "Radu Grosu" ]
We introduce a new class of time-continuous recurrent neural network models. Instead of declaring a learning system's dynamics by implicit nonlinearities, we construct networks of linear first-order dynamical systems modulated via nonlinear interlinked gates. The resulting models represent dynamical systems with varyin...
main
Machine Learning
10.1609/aaai.v35i9.16936
35
9
7657-7666
official
2006.04439
title_snapshot
10.1609/aaai.v35i9.16937
Learning with Safety Constraints: Sample Complexity of Reinforcement Learning for Constrained MDPs
https://ojs.aaai.org/index.php/AAAI/article/view/16937
https://ojs.aaai.org/index.php/AAAI/article/download/16937/16744
[ "Aria HasanzadeZonuzy", "Archana Bura", "Dileep Kalathil", "Srinivas Shakkottai" ]
Many physical systems have underlying safety considerations that require that the policy employed ensures the satisfaction of a set of constraints. The analytical formulation usually takes the form of a Constrained Markov Decision Process (CMDP). We focus on the case where the CMDP is unknown, and RL algorithms obtain ...
main
Machine Learning
10.1609/aaai.v35i9.16937
35
9
7667-7674
official
2008.00311
title_snapshot
10.1609/aaai.v35i9.16938
Analysing the Noise Model Error for Realistic Noisy Label Data
https://ojs.aaai.org/index.php/AAAI/article/view/16938
https://ojs.aaai.org/index.php/AAAI/article/download/16938/16745
[ "Michael A. Hedderich", "Dawei Zhu", "Dietrich Klakow" ]
Distant and weak supervision allow to obtain large amounts of labeled training data quickly and cheaply, but these automatic annotations tend to contain a high amount of errors. A popular technique to overcome the negative effects of these noisy labels is noise modelling where the underlying noise process is modelled. ...
main
Machine Learning
10.1609/aaai.v35i9.16938
35
9
7675-7684
official
2101.09763
title_snapshot
10.1609/aaai.v35i9.16939
Provably Good Solutions to the Knapsack Problem via Neural Networks of Bounded Size
https://ojs.aaai.org/index.php/AAAI/article/view/16939
https://ojs.aaai.org/index.php/AAAI/article/download/16939/16746
[ "Christoph Hertrich", "Martin Skutella" ]
The development of a satisfying and rigorous mathematical understanding of the performance of neural networks is a major challenge in artificial intelligence. Against this background, we study the expressive power of neural networks through the example of the classical NP-hard Knapsack Problem. Our main contribution is...
main
Machine Learning
10.1609/aaai.v35i9.16939
35
9
7685-7693
official
2005.14105
title_snapshot
10.1609/aaai.v35i9.16940
Scaling-Up Robust Gradient Descent Techniques
https://ojs.aaai.org/index.php/AAAI/article/view/16940
https://ojs.aaai.org/index.php/AAAI/article/download/16940/16747
[ "Matthew J. Holland" ]
We study a scalable alternative to robust gradient descent (RGD) techniques that can be used when losses and/or gradients can be heavy-tailed, though this will be unknown to the learner. The core technique is simple: instead of trying to robustly aggregate gradients at each step, which is costly and leads to sub-optima...
main
Machine Learning
10.1609/aaai.v35i9.16940
35
9
7694-7701
official
null
null
10.1609/aaai.v35i9.16941
Learning Model-Based Privacy Protection under Budget Constraints
https://ojs.aaai.org/index.php/AAAI/article/view/16941
https://ojs.aaai.org/index.php/AAAI/article/download/16941/16748
[ "Junyuan Hong", "Haotao Wang", "Zhangyang Wang", "Jiayu Zhou" ]
Protecting privacy in gradient-based learning has become increasingly critical as more sensitive information is being used. Many existing solutions seek to protect the sensitive gradients by constraining the overall privacy cost within a constant budget, where the protection is hand-designed and empirically calibrated ...
main
Machine Learning
10.1609/aaai.v35i9.16941
35
9
7702-7710
official
null
null
10.1609/aaai.v35i9.16942
Graph Game Embedding
https://ojs.aaai.org/index.php/AAAI/article/view/16942
https://ojs.aaai.org/index.php/AAAI/article/download/16942/16749
[ "Xiaobin Hong", "Tong Zhang", "Zhen Cui", "Yuge Huang", "Pengcheng Shen", "Shaoxin Li", "Jian Yang" ]
Graph embedding aims to encode nodes/edges into low-dimensional continuous features, and has become a crucial tool for graph analysis including graph/node classification, link prediction, etc. In this paper we propose a novel graph learning framework, named graph game embedding, to learn discriminative node representat...
main
Machine Learning
10.1609/aaai.v35i9.16942
35
9
7711-7720
official
null
null
10.1609/aaai.v35i9.16917
HiGAN: Handwriting Imitation Conditioned on Arbitrary-Length Texts and Disentangled Styles
https://ojs.aaai.org/index.php/AAAI/article/view/16917
https://ojs.aaai.org/index.php/AAAI/article/download/16917/16724
[ "Ji Gan", "Weiqiang Wang" ]
Given limited handwriting scripts, humans can easily visualize (or imagine) what the handwritten words/texts would look like with other arbitrary textual contents. Moreover, a person also is able to imitate the handwriting styles of provided reference samples. Humans can do such hallucinations, perhaps because they can...
main
Machine Learning
10.1609/aaai.v35i9.16917
35
9
7484-7492
official
null
null
10.1609/aaai.v35i9.16918
Diffusion Network Inference from Partial Observations
https://ojs.aaai.org/index.php/AAAI/article/view/16918
https://ojs.aaai.org/index.php/AAAI/article/download/16918/16725
[ "Ting Gan", "Keqi Han", "Hao Huang", "Shi Ying", "Yunjun Gao", "Zongpeng Li" ]
To infer the structure of a diffusion network from observed diffusion results, existing approaches customarily assume that observed data are complete and contain the final infection status of each node, as well as precise timestamps of node infections. Due to high cost and uncertainties in the monitoring of node infect...
main
Machine Learning
10.1609/aaai.v35i9.16918
35
9
7493-7500
official
null
null
10.1609/aaai.v35i9.16919
Stabilizing Q Learning Via Soft Mellowmax Operator
https://ojs.aaai.org/index.php/AAAI/article/view/16919
https://ojs.aaai.org/index.php/AAAI/article/download/16919/16726
[ "Yaozhong Gan", "Zhe Zhang", "Xiaoyang Tan" ]
Learning complicated value functions in high dimensional state space by function approximation is a challenging task, partially due to that the max-operator used in temporal difference updates can theoretically cause instability for most linear or non-linear approximation schemes. Mellowmax is a recently proposed diffe...
main
Machine Learning
10.1609/aaai.v35i9.16919
35
9
7501-7509
official
2012.09456
title_snapshot
10.1609/aaai.v35i9.16920
On the Convergence of Communication-Efficient Local SGD for Federated Learning
https://ojs.aaai.org/index.php/AAAI/article/view/16920
https://ojs.aaai.org/index.php/AAAI/article/download/16920/16727
[ "Hongchang Gao", "An Xu", "Heng Huang" ]
Federated Learning (FL) has attracted increasing attention in recent years. A leading training algorithm in FL is local SGD, which updates the model parameter on each worker and averages model parameters across different workers only once in a while. Although it has fewer communication rounds than the classical paralle...
main
Machine Learning
10.1609/aaai.v35i9.16920
35
9
7510-7518
official
null
null
10.1609/aaai.v35i9.16921
A Trace-restricted Kronecker-Factored Approximation to Natural Gradient
https://ojs.aaai.org/index.php/AAAI/article/view/16921
https://ojs.aaai.org/index.php/AAAI/article/download/16921/16728
[ "Kaixin Gao", "Xiaolei Liu", "Zhenghai Huang", "Min Wang", "Zidong Wang", "Dachuan Xu", "Fan Yu" ]
Second-order optimization methods have the ability to accelerate convergence by modifying the gradient through the curvature matrix. There have been many attempts to use second-order optimization methods for training deep neural networks. In this work, inspired by diagonal approximations and factored approximations suc...
main
Machine Learning
10.1609/aaai.v35i9.16921
35
9
7519-7527
official
2011.10741
title_snapshot
10.1609/aaai.v35i9.16922
Addressing Domain Gap via Content Invariant Representation for Semantic Segmentation
https://ojs.aaai.org/index.php/AAAI/article/view/16922
https://ojs.aaai.org/index.php/AAAI/article/download/16922/16729
[ "Li Gao", "Lefei Zhang", "Qian Zhang" ]
The problem of unsupervised domain adaptation in semantic segmentation is a major challenge for numerous computer vision tasks because acquiring pixel-level labels is time-consuming with expensive human labor. A large gap exists among data distributions in different domains, which will cause severe performance loss whe...
main
Machine Learning
10.1609/aaai.v35i9.16922
35
9
7528-7536
official
null
null
10.1609/aaai.v35i10.17123
Tempered Sigmoid Activations for Deep Learning with Differential Privacy
https://ojs.aaai.org/index.php/AAAI/article/view/17123
https://ojs.aaai.org/index.php/AAAI/article/download/17123/16930
[ "Nicolas Papernot", "Abhradeep Thakurta", "Shuang Song", "Steve Chien", "Úlfar Erlingsson" ]
Because learning sometimes involves sensitive data, machine learning algorithms have been extended to offer differential privacy for training data. In practice, this has been mostly an afterthought, with privacy-preserving models obtained by re-running training with a different optimizer, but using the model architectu...
main
Machine Learning
10.1609/aaai.v35i10.17123
35
10
9312-9321
official
2007.14191
title_snapshot
10.1609/aaai.v35i10.17124
Vector Quantized Bayesian Neural Network Inference for Data Streams
https://ojs.aaai.org/index.php/AAAI/article/view/17124
https://ojs.aaai.org/index.php/AAAI/article/download/17124/16931
[ "Namuk Park", "Taekyu Lee", "Songkuk Kim" ]
Bayesian neural networks (BNN) can estimate the uncertainty in predictions, as opposed to non-Bayesian neural networks (NNs). However, BNNs have been far less widely used than non-Bayesian NNs in practice since they need iterative NN executions to predict a result for one data, and it gives rise to prohibitive computat...
main
Machine Learning
10.1609/aaai.v35i10.17124
35
10
9322-9330
official
1907.05911
title_snapshot
10.1609/aaai.v35i10.17125
Maximum Roaming Multi-Task Learning
https://ojs.aaai.org/index.php/AAAI/article/view/17125
https://ojs.aaai.org/index.php/AAAI/article/download/17125/16932
[ "Lucas Pascal", "Pietro Michiardi", "Xavier Bost", "Benoit Huet", "Maria A. Zuluaga" ]
Multi-task learning has gained popularity due to the advantages it provides with respect to resource usage and performance. Nonetheless, the joint optimization of parameters with respect to multiple tasks remains an active research topic. Sub-partitioning the parameters between different tasks has proven to be an effic...
main
Machine Learning
10.1609/aaai.v35i10.17125
35
10
9331-9341
official
2006.09762
title_snapshot
10.1609/aaai.v35i10.17126
Fast PCA in 1-D Wasserstein Spaces via B-splines Representation and Metric Projection
https://ojs.aaai.org/index.php/AAAI/article/view/17126
https://ojs.aaai.org/index.php/AAAI/article/download/17126/16933
[ "Matteo Pegoraro", "Mario Beraha" ]
We address the problem of performing Principal Component Analysis over a family of probability measures on the real line, using the Wasserstein geometry. We present a novel representation of the 2-Wasserstein space, based on a well known isometric bijection and a B-spline expansion. Thanks to this representation, we ar...
main
Machine Learning
10.1609/aaai.v35i10.17126
35
10
9342-9349
official
null
null
10.1609/aaai.v35i10.17103
Top-k Ranking Bayesian Optimization
https://ojs.aaai.org/index.php/AAAI/article/view/17103
https://ojs.aaai.org/index.php/AAAI/article/download/17103/16910
[ "Quoc Phong Nguyen", "Sebastian Tay", "Bryan Kian Hsiang Low", "Patrick Jaillet" ]
This paper presents a novel approach to top-k ranking Bayesian optimization (top-k ranking BO) which is a practical and significant generalization of preferential BO to handle top-k ranking and tie/indifference observations. We first design a surrogate model that is not only capable of catering to the above observation...
main
Machine Learning
10.1609/aaai.v35i10.17103
35
10
9135-9143
official
2012.10688
title_snapshot
10.1609/aaai.v35i10.17104
Distributional Reinforcement Learning via Moment Matching
https://ojs.aaai.org/index.php/AAAI/article/view/17104
https://ojs.aaai.org/index.php/AAAI/article/download/17104/16911
[ "Thanh Nguyen-Tang", "Sunil Gupta", "Svetha Venkatesh" ]
We consider the problem of learning a set of probability distributions from the empirical Bellman dynamics in distributional reinforcement learning (RL), a class of state-of-the-art methods that estimate the distribution, as opposed to only the expectation, of the total return. We formulate a method that learns a finit...
main
Machine Learning
10.1609/aaai.v35i10.17104
35
10
9144-9152
official
2007.12354
title_snapshot
10.1609/aaai.v35i10.17105
Precision-based Boosting
https://ojs.aaai.org/index.php/AAAI/article/view/17105
https://ojs.aaai.org/index.php/AAAI/article/download/17105/16912
[ "Mohammad Hossein Nikravan", "Marjan Movahedan", "Sandra Zilles" ]
AdaBoost is a highly popular ensemble classification method for which many variants have been published. This paper proposes a generic refinement of all of these AdaBoost variants. Instead of assigning weights based on the total error of the base classifiers (as in AdaBoost), our method uses class-specific error rates....
main
Machine Learning
10.1609/aaai.v35i10.17105
35
10
9153-9160
official
null
null
10.1609/aaai.v35i10.17106
Improving Model Robustness by Adaptively Correcting Perturbation Levels with Active Queries
https://ojs.aaai.org/index.php/AAAI/article/view/17106
https://ojs.aaai.org/index.php/AAAI/article/download/17106/16913
[ "Kun-Peng Ning", "Lue Tao", "Songcan Chen", "Sheng-Jun Huang" ]
In addition to high accuracy, robustness is becoming increasingly important for machine learning models in various applications. Recently, much research has been devoted to improving the model robustness by training with noise perturbations. Most existing studies assume a fixed perturbation level for all training examp...
main
Machine Learning
10.1609/aaai.v35i10.17106
35
10
9161-9169
official
2103.14824
title_snapshot
10.1609/aaai.v35i10.17107
Learning of Structurally Unambiguous Probabilistic Grammars
https://ojs.aaai.org/index.php/AAAI/article/view/17107
https://ojs.aaai.org/index.php/AAAI/article/download/17107/16914
[ "Dolav Nitay", "Dana Fisman", "Michal Ziv-Ukelson" ]
The problem of identifying a probabilistic context free grammar has two aspects: the first is determining the grammar's topology (the rules of the grammar) and the second is estimating probabilistic weights for each rule. Given the hardness results for learning context-free grammars in general, and probabilistic gramma...
main
Machine Learning
10.1609/aaai.v35i10.17107
35
10
9170-9178
official
2011.07472
title_snapshot
10.1609/aaai.v35i10.17108
RT3D: Achieving Real-Time Execution of 3D Convolutional Neural Networks on Mobile Devices
https://ojs.aaai.org/index.php/AAAI/article/view/17108
https://ojs.aaai.org/index.php/AAAI/article/download/17108/16915
[ "Wei Niu", "Mengshu Sun", "Zhengang Li", "Jou-An Chen", "Jiexiong Guan", "Xipeng Shen", "Yanzhi Wang", "Sijia Liu", "Xue Lin", "Bin Ren" ]
Mobile devices are becoming an important carrier for deep learning tasks, as they are being equipped with powerful, high-end mobile CPUs and GPUs. However, it is still a challenging task to execute 3D Convolutional Neural Networks (CNNs) targeting for real-time performance, besides high inference accuracy. The reason i...
main
Machine Learning
10.1609/aaai.v35i10.17108
35
10
9179-9187
official
2007.09835
title_snapshot
10.1609/aaai.v35i10.17109
Warm Starting CMA-ES for Hyperparameter Optimization
https://ojs.aaai.org/index.php/AAAI/article/view/17109
https://ojs.aaai.org/index.php/AAAI/article/download/17109/16916
[ "Masahiro Nomura", "Shuhei Watanabe", "Youhei Akimoto", "Yoshihiko Ozaki", "Masaki Onishi" ]
Hyperparameter optimization (HPO), formulated as black-box optimization (BBO), is recognized as essential for automation and high performance of machine learning approaches. The CMA-ES is a promising BBO approach with a high degree of parallelism, and has been applied to HPO tasks, often under parallel implementation, ...
main
Machine Learning
10.1609/aaai.v35i10.17109
35
10
9188-9196
official
2012.06932
title_snapshot
10.1609/aaai.v35i10.17110
Inverse Reinforcement Learning From Like-Minded Teachers
https://ojs.aaai.org/index.php/AAAI/article/view/17110
https://ojs.aaai.org/index.php/AAAI/article/download/17110/16917
[ "Ritesh Noothigattu", "Tom Yan", "Ariel D. Procaccia" ]
We study the problem of learning a policy in a Markov decision process (MDP) based on observations of the actions taken by multiple teachers. We assume that the teachers are like-minded in that their reward functions -- while different from each other -- are random perturbations of an underlying reward function. Under ...
main
Machine Learning
10.1609/aaai.v35i10.17110
35
10
9197-9204
official
null
null
10.1609/aaai.v35i10.17111
Multinomial Logit Contextual Bandits: Provable Optimality and Practicality
https://ojs.aaai.org/index.php/AAAI/article/view/17111
https://ojs.aaai.org/index.php/AAAI/article/download/17111/16918
[ "Min-hwan Oh", "Garud Iyengar" ]
We consider a sequential assortment selection problem where the user choice is given by a multinomial logit (MNL) choice model whose parameters are unknown. In each period, the learning agent observes a d-dimensional contextual information about the user and the N available items, and offers an assortment of size K to ...
main
Machine Learning
10.1609/aaai.v35i10.17111
35
10
9205-9213
official
2103.13929
title_snapshot
10.1609/aaai.v35i10.17112
Learning Deep Generative Models for Queuing Systems
https://ojs.aaai.org/index.php/AAAI/article/view/17112
https://ojs.aaai.org/index.php/AAAI/article/download/17112/16919
[ "Cesar Ojeda", "Kostadin Cvejoski", "Bodgan Georgiev", "Christian Bauckhage", "Jannis Schuecker", "Ramses J. Sanchez" ]
Modern society is heavily dependent on large scale client-server systems with applications ranging from Internet and Communication Services to sophisticated logistics and deployment of goods. To maintain and improve such a system, a careful study of client and server dynamics is needed – e.g. response/service times, av...
main
Machine Learning
10.1609/aaai.v35i10.17112
35
10
9214-9222
official
null
null
10.1609/aaai.v35i10.17113
OT-Flow: Fast and Accurate Continuous Normalizing Flows via Optimal Transport
https://ojs.aaai.org/index.php/AAAI/article/view/17113
https://ojs.aaai.org/index.php/AAAI/article/download/17113/16920
[ "Derek Onken", "Samy Wu Fung", "Xingjian Li", "Lars Ruthotto" ]
A normalizing flow is an invertible mapping between an arbitrary probability distribution and a standard normal distribution; it can be used for density estimation and statistical inference. Computing the flow follows the change of variables formula and thus requires invertibility of the mapping and an efficient way to...
main
Machine Learning
10.1609/aaai.v35i10.17113
35
10
9223-9232
official
2006.00104
title_snapshot
10.1609/aaai.v35i10.17114
FC-GAGA: Fully Connected Gated Graph Architecture for Spatio-Temporal Traffic Forecasting
https://ojs.aaai.org/index.php/AAAI/article/view/17114
https://ojs.aaai.org/index.php/AAAI/article/download/17114/16921
[ "Boris N. Oreshkin", "Arezou Amini", "Lucy Coyle", "Mark Coates" ]
Forecasting of multivariate time-series is an important problem that has applications in traffic management, cellular network configuration, and quantitative finance. A special case of the problem arises when there is a graph available that captures the relationships between the time-series. In this paper we propose a ...
main
Machine Learning
10.1609/aaai.v35i10.17114
35
10
9233-9241
official
2007.15531
title_snapshot
10.1609/aaai.v35i10.17115
Meta-Learning Framework with Applications to Zero-Shot Time-Series Forecasting
https://ojs.aaai.org/index.php/AAAI/article/view/17115
https://ojs.aaai.org/index.php/AAAI/article/download/17115/16922
[ "Boris N. Oreshkin", "Dmitri Carpov", "Nicolas Chapados", "Yoshua Bengio" ]
Can meta-learning discover generic ways of processing time series (TS) from a diverse dataset so as to greatly improve generalization on new TS coming from different datasets? This work provides positive evidence to this using a broad meta-learning framework which we show subsumes many existing meta-learning algorithms...
main
Machine Learning
10.1609/aaai.v35i10.17115
35
10
9242-9250
official
2002.02887
title_snapshot
10.1609/aaai.v35i10.17116
Augmented Experiment in Material Engineering Using Machine Learning
https://ojs.aaai.org/index.php/AAAI/article/view/17116
https://ojs.aaai.org/index.php/AAAI/article/download/17116/16923
[ "Aomar Osmani", "Massinissa Hamidi", "Salah Bouhouche" ]
The synthesis of materials using the principle of thermogravimetric analysis to discover new anticorrosive paints requires several costly experiments. This paper presents an approach combining empirical data and domain analytical models to reduce the number of real experiments required to obtain the desired synthesis. ...
main
Machine Learning
10.1609/aaai.v35i10.17116
35
10
9251-9258
official
null
null
10.1609/aaai.v35i10.17117
Second Order Techniques for Learning Time-series with Structural Breaks
https://ojs.aaai.org/index.php/AAAI/article/view/17117
https://ojs.aaai.org/index.php/AAAI/article/download/17117/16924
[ "Takayuki Osogami" ]
We study fundamental problems in learning nonstationary time-series: how to effectively regularize time-series models and how to adaptively tune forgetting rates. The effectiveness of L2 regularization depends on the choice of coordinates, and the variables need to be appropriately normalized. In nonstationary environm...
main
Machine Learning
10.1609/aaai.v35i10.17117
35
10
9259-9267
official
null
null
10.1609/aaai.v35i10.17118
Defending against Backdoors in Federated Learning with Robust Learning Rate
https://ojs.aaai.org/index.php/AAAI/article/view/17118
https://ojs.aaai.org/index.php/AAAI/article/download/17118/16925
[ "Mustafa Safa Ozdayi", "Murat Kantarcioglu", "Yulia R. Gel" ]
Federated learning (FL) allows a set of agents to collaboratively train a model without sharing their potentially sensitive data. This makes FL suitable for privacy-preserving applications. At the same time, FL is susceptible to adversarial attacks due to decentralized and unvetted data. One important line of attacks a...
main
Machine Learning
10.1609/aaai.v35i10.17118
35
10
9268-9276
official
2007.03767
title_snapshot
10.1609/aaai.v35i10.17119
Robustness Guarantees for Mode Estimation with an Application to Bandits
https://ojs.aaai.org/index.php/AAAI/article/view/17119
https://ojs.aaai.org/index.php/AAAI/article/download/17119/16926
[ "Aldo Pacchiano", "Heinrich Jiang", "Michael I. Jordan" ]
Mode estimation is a classical problem in statistics with a wide range of applications in machine learning. Despite this, there is little understanding in its robustness properties under possibly adversarial data contamination. In this paper, we give precise robustness guarantees as well as privacy guarantees under sim...
main
Machine Learning
10.1609/aaai.v35i10.17119
35
10
9277-9284
official
2003.02932
title_snapshot
10.1609/aaai.v35i10.17120
Disentangled Information Bottleneck
https://ojs.aaai.org/index.php/AAAI/article/view/17120
https://ojs.aaai.org/index.php/AAAI/article/download/17120/16927
[ "Ziqi Pan", "Li Niu", "Jianfu Zhang", "Liqing Zhang" ]
The information bottleneck (IB) method is a technique for extracting information that is relevant for predicting the target random variable from the source random variable, which is typically implemented by optimizing the IB Lagrangian that balances the compression and prediction terms. However, the IB Lagrangian is ha...
main
Machine Learning
10.1609/aaai.v35i10.17120
35
10
9285-9293
official
2012.07372
title_snapshot
10.1609/aaai.v35i10.17121
NASTransfer: Analyzing Architecture Transferability in Large Scale Neural Architecture Search
https://ojs.aaai.org/index.php/AAAI/article/view/17121
https://ojs.aaai.org/index.php/AAAI/article/download/17121/16928
[ "Rameswar Panda", "Michele Merler", "Mayoore S Jaiswal", "Hui Wu", "Kandan Ramakrishnan", "Ulrich Finkler", "Chun-Fu Richard Chen", "Minsik Cho", "Rogerio Feris", "David Kung", "Bishwaranjan Bhattacharjee" ]
Neural Architecture Search (NAS) is an open and challenging problem in machine learning. While NAS offers great promise, the prohibitive computational demand of most of the existing NAS methods makes it difficult to directly search the architectures on large-scale tasks. The typical way of conducting large scale NAS is...
main
Machine Learning
10.1609/aaai.v35i10.17121
35
10
9294-9302
official
2006.13314
title_snapshot
10.1609/aaai.v35i10.17122
Robust Reinforcement Learning: A Case Study in Linear Quadratic Regulation
https://ojs.aaai.org/index.php/AAAI/article/view/17122
https://ojs.aaai.org/index.php/AAAI/article/download/17122/16929
[ "Bo Pang", "Zhong-Ping Jiang" ]
This paper studies the robustness of reinforcement learning algorithms to errors in the learning process. Specifically, we revisit the benchmark problem of discrete-time linear quadratic regulation (LQR) and study the long-standing open question: Under what conditions is the policy iteration method robustly stable from...
main
Machine Learning
10.1609/aaai.v35i10.17122
35
10
9303-9311
official
2008.11592
title_snapshot
10.1609/aaai.v35i10.17083
Policy Optimization as Online Learning with Mediator Feedback
https://ojs.aaai.org/index.php/AAAI/article/view/17083
https://ojs.aaai.org/index.php/AAAI/article/download/17083/16890
[ "Alberto Maria Metelli", "Matteo Papini", "Pierluca D'Oro", "Marcello Restelli" ]
Policy Optimization (PO) is a widely used approach to address continuous control tasks. In this paper, we introduce the notion of mediator feedback that frames PO as an online learning problem over the policy space. The additional available information, compared to the standard bandit feedback, allows reusing samples g...
main
Machine Learning
10.1609/aaai.v35i10.17083
35
10
8958-8966
official
2012.08225
title_snapshot
10.1609/aaai.v35i10.17084
Consistency and Finite Sample Behavior of Binary Class Probability Estimation
https://ojs.aaai.org/index.php/AAAI/article/view/17084
https://ojs.aaai.org/index.php/AAAI/article/download/17084/16891
[ "Alexander Mey", "Marco Loog" ]
We investigate to which extent one can recover class probabilities within the empirical risk minimization (ERM) paradigm. We extend existing results and emphasize the tight relations between empirical risk minimization and class probability estimation. Following previous literature on excess risk bounds and proper scor...
main
Machine Learning
10.1609/aaai.v35i10.17084
35
10
8967-8974
official
1908.11823
title_snapshot
10.1609/aaai.v35i10.17085
Discovering Fully Oriented Causal Networks
https://ojs.aaai.org/index.php/AAAI/article/view/17085
https://ojs.aaai.org/index.php/AAAI/article/download/17085/16892
[ "Osman A Mian", "Alexander Marx", "Jilles Vreeken" ]
We study the problem of inferring causal graphs from observational data. We are particularly interested in discovering graphs where all edges are oriented, as opposed to the partially directed graph that the state of the art discover. To this end, we base our approach on the algorithmic Markov condition. Unlike the sta...
main
Machine Learning
10.1609/aaai.v35i10.17085
35
10
8975-8982
official
null
null
10.1609/aaai.v35i10.17086
Generative Semi-supervised Learning for Multivariate Time Series Imputation
https://ojs.aaai.org/index.php/AAAI/article/view/17086
https://ojs.aaai.org/index.php/AAAI/article/download/17086/16893
[ "Xiaoye Miao", "Yangyang Wu", "Jun Wang", "Yunjun Gao", "Xudong Mao", "Jianwei Yin" ]
The missing values, widely existed in multivariate time series data, hinder the effective data analysis. Existing time series imputation methods do not make full use of the label information in real-life time series data. In this paper, we propose a novel semi-supervised generative adversarial network model, named SSGA...
main
Machine Learning
10.1609/aaai.v35i10.17086
35
10
8983-8991
official
null
null
10.1609/aaai.v35i10.17087
A General Class of Transfer Learning Regression without Implementation Cost
https://ojs.aaai.org/index.php/AAAI/article/view/17087
https://ojs.aaai.org/index.php/AAAI/article/download/17087/16894
[ "Shunya Minami", "Song Liu", "Stephen Wu", "Kenji Fukumizu", "Ryo Yoshida" ]
We propose a novel framework that unifies and extends existing methods of transfer learning (TL) for regression. To bridge a pretrained source model to the model on a target task, we introduce a density-ratio reweighting function, which is estimated through the Bayesian framework with a specific prior distribution. By ...
main
Machine Learning
10.1609/aaai.v35i10.17087
35
10
8992-8999
official
2006.13228
title_snapshot
10.1609/aaai.v35i10.17088
Scheduling of Time-Varying Workloads Using Reinforcement Learning
https://ojs.aaai.org/index.php/AAAI/article/view/17088
https://ojs.aaai.org/index.php/AAAI/article/download/17088/16895
[ "Shanka Subhra Mondal", "Nikhil Sheoran", "Subrata Mitra" ]
Resource usage of production workloads running on shared compute clusters often fluctuate significantly across time. While simultaneous spike in the resource usage between two workloads running on the same machine can create performance degradation, unused resources in a machine results in wastage and undesirable opera...
main
Machine Learning
10.1609/aaai.v35i10.17088
35
10
9000-9008
official
null
null
10.1609/aaai.v35i10.17089
Improved Mutual Information Estimation
https://ojs.aaai.org/index.php/AAAI/article/view/17089
https://ojs.aaai.org/index.php/AAAI/article/download/17089/16896
[ "Youssef Mroueh", "Igor Melnyk", "Pierre Dognin", "Jarret Ross", "Tom Sercu" ]
We propose to estimate the KL divergence using a relaxed likelihood ratio estimation in a Reproducing Kernel Hilbert space. We show that the dual of our ratio estimator for KL in the particular case of Mutual Information estimation corresponds to a lower bound on the MI that is related to the so called Donsker Varadhan...
main
Machine Learning
10.1609/aaai.v35i10.17089
35
10
9009-9017
official
null
null
10.1609/aaai.v35i10.17090
Text-based RL Agents with Commonsense Knowledge: New Challenges, Environments and Baselines
https://ojs.aaai.org/index.php/AAAI/article/view/17090
https://ojs.aaai.org/index.php/AAAI/article/download/17090/16897
[ "Keerthiram Murugesan", "Mattia Atzeni", "Pavan Kapanipathi", "Pushkar Shukla", "Sadhana Kumaravel", "Gerald Tesauro", "Kartik Talamadupula", "Mrinmaya Sachan", "Murray Campbell" ]
Text-based games have emerged as an important test-bed for Reinforcement Learning (RL) research, requiring RL agents to combine grounded language understanding with sequential decision making. In this paper, we examine the problem of infusing RL agents with commonsense knowledge. Such knowledge would allow agents to ef...
main
Machine Learning
10.1609/aaai.v35i10.17090
35
10
9018-9027
official
2010.03790
title_snapshot
10.1609/aaai.v35i10.17091
Task-Agnostic Exploration via Policy Gradient of a Non-Parametric State Entropy Estimate
https://ojs.aaai.org/index.php/AAAI/article/view/17091
https://ojs.aaai.org/index.php/AAAI/article/download/17091/16898
[ "Mirco Mutti", "Lorenzo Pratissoli", "Marcello Restelli" ]
In a reward-free environment, what is a suitable intrinsic objective for an agent to pursue so that it can learn an optimal task-agnostic exploration policy? In this paper, we argue that the entropy of the state distribution induced by finite-horizon trajectories is a sensible target. Especially, we present a novel and...
main
Machine Learning
10.1609/aaai.v35i10.17091
35
10
9028-9036
official
2007.04640
title_snapshot
10.1609/aaai.v35i10.17092
Elastic Consistency: A Practical Consistency Model for Distributed Stochastic Gradient Descent
https://ojs.aaai.org/index.php/AAAI/article/view/17092
https://ojs.aaai.org/index.php/AAAI/article/download/17092/16899
[ "Giorgi Nadiradze", "Ilia Markov", "Bapi Chatterjee", "Vyacheslav Kungurtsev", "Dan Alistarh" ]
One key element behind the recent progress of machine learning has been the ability to train machine learning models in large-scale distributed shared-memory and message-passing environments. Most of these models are trained employing variants of stochastic gradient descent (SGD) based optimization, but most methods in...
main
Machine Learning
10.1609/aaai.v35i10.17092
35
10
9037-9045
official
2001.05918
title_judge
10.1609/aaai.v35i10.17093
Game of Gradients: Mitigating Irrelevant Clients in Federated Learning
https://ojs.aaai.org/index.php/AAAI/article/view/17093
https://ojs.aaai.org/index.php/AAAI/article/download/17093/16900
[ "Lokesh Nagalapatti", "Ramasuri Narayanam" ]
The paradigm of Federated learning (FL) deals with multiple clients participating in collaborative training of a machine learning model under the orchestration of a central server. In this setup, each client’s data is private to itself and is not transferable to other clients or the server. Though FL paradigm has recei...
main
Machine Learning
10.1609/aaai.v35i10.17093
35
10
9046-9054
official
2110.12257
title_snapshot
10.1609/aaai.v35i10.17094
Objective-Based Hierarchical Clustering of Deep Embedding Vectors
https://ojs.aaai.org/index.php/AAAI/article/view/17094
https://ojs.aaai.org/index.php/AAAI/article/download/17094/16901
[ "Stanislav Naumov", "Grigory Yaroslavtsev", "Dmitrii Avdiukhin" ]
We initiate a comprehensive experimental study of objective-based hierarchical clustering methods on massive datasets consisting of deep embedding vectors from computer vision and NLP applications. This includes a large variety of image embedding (ImageNet, ImageNetV2, NaBirds), word embedding (Twitter, Wikipedia), and...
main
Machine Learning
10.1609/aaai.v35i10.17094
35
10
9055-9063
official
2012.08466
title_snapshot
10.1609/aaai.v35i10.17095
5* Knowledge Graph Embeddings with Projective Transformations
https://ojs.aaai.org/index.php/AAAI/article/view/17095
https://ojs.aaai.org/index.php/AAAI/article/download/17095/16902
[ "Mojtaba Nayyeri", "Sahar Vahdati", "Can Aykul", "Jens Lehmann" ]
Performing link prediction using knowledge graph embedding models has become a popular approach for knowledge graph completion. Such models employ a transformation function that maps nodes via edges into a vector space in order to measure the likelihood of the links. While mapping the individual nodes, the structure of...
main
Machine Learning
10.1609/aaai.v35i10.17095
35
10
9064-9072
official
2006.04986
title_snapshot
10.1609/aaai.v35i10.17096
Advice-Guided Reinforcement Learning in a non-Markovian Environment
https://ojs.aaai.org/index.php/AAAI/article/view/17096
https://ojs.aaai.org/index.php/AAAI/article/download/17096/16903
[ "Daniel Neider", "Jean-Raphael Gaglione", "Ivan Gavran", "Ufuk Topcu", "Bo Wu", "Zhe Xu" ]
We study a class of reinforcement learning tasks in which the agent receives its reward for complex, temporally-extended behaviors sparsely. For such tasks, the problem is how to augment the state-space so as to make the reward function Markovian in an efficient way. While some existing solutions assume that the reward...
main
Machine Learning
10.1609/aaai.v35i10.17096
35
10
9073-9080
official
null
null
10.1609/aaai.v35i10.17097
Clinical Risk Prediction with Temporal Probabilistic Asymmetric Multi-Task Learning
https://ojs.aaai.org/index.php/AAAI/article/view/17097
https://ojs.aaai.org/index.php/AAAI/article/download/17097/16904
[ "A. Tuan Nguyen", "Hyewon Jeong", "Eunho Yang", "Sung Ju Hwang" ]
Although recent multi-task learning methods have shown to be effective in improving the generalization of deep neural networks, they should be used with caution for safety-critical applications, such as clinical risk prediction. This is because even if they achieve improved task-average performance, they may still yiel...
main
Machine Learning
10.1609/aaai.v35i10.17097
35
10
9081-9091
official
2006.12777
title_snapshot
10.1609/aaai.v35i10.17098
Modular Graph Transformer Networks for Multi-Label Image Classification
https://ojs.aaai.org/index.php/AAAI/article/view/17098
https://ojs.aaai.org/index.php/AAAI/article/download/17098/16905
[ "Hoang D. Nguyen", "Xuan-Son Vu", "Duc-Trong Le" ]
With the recent advances in graph neural networks, there is a rising number of studies on graph-based multi-label classification with the consideration of object dependencies within visual data. Nevertheless, graph representations can become indistinguishable due to the complex nature of label relationships. We propose...
main
Machine Learning
10.1609/aaai.v35i10.17098
35
10
9092-9100
official
null
null
10.1609/aaai.v35i10.17099
Differentially Private k-Means via Exponential Mechanism and Max Cover
https://ojs.aaai.org/index.php/AAAI/article/view/17099
https://ojs.aaai.org/index.php/AAAI/article/download/17099/16906
[ "Huy L. Nguyen", "Anamay Chaturvedi", "Eric Z Xu" ]
We introduce a new (ϵₚ, δₚ)-differentially private algorithm for the k-means clustering problem. Given a dataset in Euclidean space, the k-means clustering problem requires one to find k points in that space such that the sum of squares of Euclidean distances between each data point and its closest respective point amo...
main
Machine Learning
10.1609/aaai.v35i10.17099
35
10
9101-9108
official
2009.01220
title_judge
10.1609/aaai.v35i10.17100
Minimum Robust Multi-Submodular Cover for Fairness
https://ojs.aaai.org/index.php/AAAI/article/view/17100
https://ojs.aaai.org/index.php/AAAI/article/download/17100/16907
[ "Lan N. Nguyen", "My T. Thai" ]
In this paper, we study a novel problem, Minimum Robust Multi-Submodular Cover for Fairness (MinRF), as follows: given a ground set V; m monotone submodular functions f_1,...,f_m; m thresholds T_1,...,T_m and a non-negative integer r; MinRF asks for the smallest set S such that f_i(S \ X) ≥ T_i for all i ∈ [m] and |X| ...
main
Machine Learning
10.1609/aaai.v35i10.17100
35
10
9109-9116
official
2012.07936
title_snapshot
10.1609/aaai.v35i10.17101
Temporal Latent Auto-Encoder: A Method for Probabilistic Multivariate Time Series Forecasting
https://ojs.aaai.org/index.php/AAAI/article/view/17101
https://ojs.aaai.org/index.php/AAAI/article/download/17101/16908
[ "Nam Nguyen", "Brian Quanz" ]
Probabilistic forecasting of high dimensional multivariate time series is a notoriously challenging task, both in terms of computational burden and distribution modeling. Most previous work either makes simple distribution assumptions or abandons modeling cross-series correlations. A promising line of work exploits sca...
main
Machine Learning
10.1609/aaai.v35i10.17101
35
10
9117-9125
official
2101.10460
title_snapshot
10.1609/aaai.v35i10.17102
An Information-Theoretic Framework for Unifying Active Learning Problems
https://ojs.aaai.org/index.php/AAAI/article/view/17102
https://ojs.aaai.org/index.php/AAAI/article/download/17102/16909
[ "Quoc Phong Nguyen", "Bryan Kian Hsiang Low", "Patrick Jaillet" ]
This paper presents an information-theoretic framework for unifying active learning problems: level set estimation (LSE), Bayesian optimization (BO), and their generalized variant. We first introduce a novel active learning criterion that subsumes an existing LSE algorithm and achieves state-of-the-art performance in L...
main
Machine Learning
10.1609/aaai.v35i10.17102
35
10
9126-9134
official
2012.10695
title_snapshot
10.1609/aaai.v35i10.17063
Tailoring Embedding Function to Heterogeneous Few-Shot Tasks by Global and Local Feature Adaptors
https://ojs.aaai.org/index.php/AAAI/article/view/17063
https://ojs.aaai.org/index.php/AAAI/article/download/17063/16870
[ "Su Lu", "Han-Jia Ye", "De-Chuan Zhan" ]
Few-Shot Learning (FSL) is essential for visual recognition. Many methods tackle this challenging problem via learning an embedding function from seen classes and transfer it to unseen classes with a few labeled instances. Researchers recently found it beneficial to incorporate task-specific feature adaptation into FSL...
main
Machine Learning
10.1609/aaai.v35i10.17063
35
10
8776-8783
official
null
null
10.1609/aaai.v35i10.17064
PULNS: Positive-Unlabeled Learning with Effective Negative Sample Selector
https://ojs.aaai.org/index.php/AAAI/article/view/17064
https://ojs.aaai.org/index.php/AAAI/article/download/17064/16871
[ "Chuan Luo", "Pu Zhao", "Chen Chen", "Bo Qiao", "Chao Du", "Hongyu Zhang", "Wei Wu", "Shaowei Cai", "Bing He", "Saravanakumar Rajmohan", "Qingwei Lin" ]
Positive-unlabeled learning (PU learning) is an important case of binary classification where the training data only contains positive and unlabeled samples. The current state-of-the-art approach for PU learning is the cost-sensitive approach, which casts PU learning as a cost-sensitive classification problem and relie...
main
Machine Learning
10.1609/aaai.v35i10.17064
35
10
8784-8792
official
null
null
10.1609/aaai.v35i10.17065
Revisiting Co-Occurring Directions: Sharper Analysis and Efficient Algorithm for Sparse Matrices
https://ojs.aaai.org/index.php/AAAI/article/view/17065
https://ojs.aaai.org/index.php/AAAI/article/download/17065/16872
[ "Luo Luo", "Cheng Chen", "Guangzeng Xie", "Haishan Ye" ]
We study the streaming model for approximate matrix multiplication (AMM). We are interested in the scenario that the algorithm can only take one pass over the data with limited memory. The state-of-the-art deterministic sketching algorithm for streaming AMM is the co-occurring directions (COD), which has much smaller a...
main
Machine Learning
10.1609/aaai.v35i10.17065
35
10
8793-8800
official
2009.02553
title_snapshot
10.1609/aaai.v35i10.17066
Semi-supervised Medical Image Segmentation through Dual-task Consistency
https://ojs.aaai.org/index.php/AAAI/article/view/17066
https://ojs.aaai.org/index.php/AAAI/article/download/17066/16873
[ "Xiangde Luo", "Jieneng Chen", "Tao Song", "Guotai Wang" ]
Deep learning-based semi-supervised learning (SSL) algorithms have led to promising results in medical images segmentation and can alleviate doctors' expensive annotations by leveraging unlabeled data. However, most of the existing SSL algorithms in literature tend to regularize the model training by perturbing network...
main
Machine Learning
10.1609/aaai.v35i10.17066
35
10
8801-8809
official
2009.04448
title_snapshot