Algorithmic Learning Theory

This book constitutes the proceedings of the 24th International Conference on Algorithmic Learning Theory, ALT 2013, held in Singapore in October 2013, and co-located with the 16th International Conference on Discovery Science, DS 2013. The 23 papers presented in this volume were carefully reviewed and selected from 39 submissions. In addition the book contains 3 full papers of invited talks. The papers are organized in topical sections named: online learning, inductive inference and grammatical inference, teaching and learning from queries, bandit theory, statistical learning theory, Bayesian/stochastic learning, and unsupervised/semi-supervised learning.

Inhalt

Editors' Introduction.- Learning and Optimizing with Preferences.- Efficient Algorithms for Combinatorial Online Prediction.- Exact Learning from Membership Queries: Some Techniques, Results and New Directions.- Online Learning Universal Algorithm for Trading in Stock Market Based on the Method of Calibration.- Combinatorial Online Prediction via Metarounding.- On Competitive Recommendations.- Online PCA with Optimal Regrets.- Inductive Inference and Grammatical Inference Partial Learning of Recursively Enumerable Languages.- Topological Separations in Inductive Inference.- PAC Learning of Some Subclasses of Context-Free Grammars with Basic Distributional Properties from Positive Data.- Universal Knowledge-Seeking Agents for Stochastic Environments.- Teaching and Learning from Queries Order Compression Schemes.- Learning a Bounded-Degree Tree Using Separator Queries.- Faster Hoeffding Racing: Bernstein Races via Jackknife Estimates.- Robust Risk-Averse Stochastic Multi-armed Bandits.- An Efficient Algorithm for Learning with Semi-bandit Feedback.- Differentially-Private Learning of Low Dimensional Manifolds.- Generalization and Robustness of Batched Weighted Average Algorithm with V-Geometrically Ergodic Markov Data.- Adaptive Metric Dimensionality Reduction.- Dimension-Adaptive Bounds on Compressive FLD Classification.- Bayesian Methods for Low-Rank Matrix Estimation: Short Survey and Theoretical Study.- Concentration and Confidence for Discrete Bayesian Sequence Predictors.- Algorithmic Connections between Active Learning and Stochastic Convex Optimization.- Unsupervised/Semi-Supervised Learning Unsupervised Model-Free Representation Learning.- Fast Spectral Clustering via the Nyström Method.- Nonparametric Multiple Change Point Estimation in Highly Dependent Time Series.