Active Learning Is Planning: Nonmyopic ε-Bayes-Optimal Active Learning of Gaussian Processes

A fundamental issue in active learning of Gaussian processes is that of the exploration-exploitation trade-off. This paper presents a novel nonmyopic e-Bayes-optimal active learning (e-BAL) approach [4] that jointly optimizes the trade-off. In contrast, existing works have primarily developed greedy algorithms or performed exploration and exploitation separately. To perform active learning in real time, we then propose an anytime algorithm [4] based on e-BAL with performance guarantee and empirically demonstrate using a real-world dataset that, with limited budget, it outperforms the state-of-the-art algorithms.

[1]  Kian Hsiang Low,et al.  Active Markov information-theoretic path planning for robotic environmental sensing , 2011, AAMAS.

[2]  Kian Hsiang Low,et al.  Decentralized active robotic exploration and mapping for probabilistic field classification in environmental sensing , 2012, AAMAS.

[3]  Kian Hsiang Low,et al.  Gaussian Process-Based Decentralized Data Fusion and Active Sensing for Mobility-on-Demand System , 2013, Robotics: Science and Systems.

[4]  Kian Hsiang Low,et al.  Information-Theoretic Approach to Efficient Adaptive Path Planning for Mobile Robotic Environmental Sensing , 2009, ICAPS.

[5]  Herbert Solomon,et al.  Optimal Design of Sampling from Finite Populations: A Critical Review and Indication of New Research Areas , 1970 .

[6]  Kian Hsiang Low,et al.  A General Framework for Interacting Bayes-Optimally with Self-Interested Agents using Arbitrary Parametric Model and Model Prior , 2013, IJCAI.

[7]  Kian Hsiang Low,et al.  Adaptive multi-robot wide-area exploration and mapping , 2008, AAMAS.

[8]  Alberto Elfes,et al.  Cooperative aquatic sensing using the telesupervised adaptive ocean sensor fleet , 2009, Remote Sensing.

[9]  Andreas Krause,et al.  Efficient Informative Sensing using Multiple Robots , 2014, J. Artif. Intell. Res..

[10]  Kian Hsiang Low,et al.  Parallel Gaussian Process Regression with Low-Rank Covariance Matrix Approximations , 2013, UAI.

[11]  Andreas Krause,et al.  Near-Optimal Sensor Placements in Gaussian Processes: Theory, Efficient Algorithms and Empirical Studies , 2008, J. Mach. Learn. Res..

[12]  Zoubin Ghahramani,et al.  Collaborative Gaussian Processes for Preference Learning , 2012, NIPS.

[13]  Peter J. Diggle,et al.  Bayesian Geostatistical Design , 2006 .

[14]  Jon Lee.  Maximum entropy sampling , 2001 .

[15]  Mijung Park,et al.  Bayesian active learning with localized priors for fast receptive field characterization , 2012, NIPS.

[16]  R. J. Martin.  Comparing and contrasting some environmental and experimental design problems , 2001 .

[17]  Kian Hsiang Low,et al.  Adaptive Sampling for Multi-Robot Wide-Area Exploration , 2007, Proceedings 2007 IEEE International Conference on Robotics and Automation.

[18]  Kian Hsiang Low,et al.  Multi-robot active sensing of non-stationary gaussian process-based environmental phenomena , 2014, AAMAS.

[19]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.

[20]  Jesse Hoey,et al.  An analytic solution to discrete Bayesian reinforcement learning , 2006, ICML.

[21]  Mohan S. Kankanhalli,et al.  Nonmyopic \(\epsilon\)-Bayes-Optimal Active Learning of Gaussian Processes , 2014, ICML.

[22]  Gaurav S. Sukhatme,et al.  Decentralized Data Fusion and Active Sensing with Mobile Sensors for Modeling and Predicting Spatiotemporal Traffic Phenomena , 2012, UAI.

[23]  Thomas M. Cover,et al.  Elements of Information Theory , 2005 .

[24]  Dale L. Zimmerman,et al.  Optimal network design for spatial prediction, covariance parameter estimation, and empirical prediction , 2006 .

[25]  Andreas Krause,et al.  Nonmyopic active learning of Gaussian processes: an exploration-exploitation approach , 2007, ICML '07.

[26]  Werner G. Müller,et al.  Collecting Spatial Data: Optimum Design of Experiments for Random Fields , 1998 .

[27]  Kian Hsiang Low,et al.  Multi-robot informative path planning for active sensing of environmental phenomena: a tale of two algorithms , 2013, AAMAS.

[28]  Kian Hsiang Low,et al.  Telesupervised remote surface water quality sensing , 2010, 2010 IEEE Aerospace Conference.