Select Lab Publications


Nonmyopic active learning of Gaussian processes: an exploration-exploitation approach (2007)

By: Andreas Krause and Carlos Guestrin

Abstract: When monitoring spatial phenomena, such as the ecological condition of a river, deciding where to make observations is a challenging task. In these settings, a fundamental ques- tion is when an active learning, or sequen- tial design, strategy, where locations are se- lected based on previous measurements, will perform significantly better than sensing at an a priori specified set of locations. For Gaussian Processes (GPs), which often accu- rately model spatial phenomena, we present an analysis and efficient algorithms that ad- dress this question. Central to our analysis is a theoretical bound which quantifies the performance difference between active and a priori design strategies. We consider GPs with unknown kernel parameters and present a nonmyopic approach for trading off ex- ploration, i.e., decreasing uncertainty about the model parameters, and exploitation, i.e., near-optimally selecting observations when the parameters are (approximately) known. We discuss several exploration strategies, and present logarithmic sample complexity bounds for the exploration phase. We then extend our algorithm to handle nonstation- ary GPs exploiting local structure in the model. We also present extensive empirical evaluation on several real-world problems.



Download Information
Andreas Krause and Carlos Guestrin (2007). "Nonmyopic active learning of Gaussian processes: an exploration-exploitation approach." International Conference on Machine Learning (ICML). pdf long talk poster  
BibTeX citation

@inproceedings{Krause+Guestrin:icml07activegp,
title = {Nonmyopic active learning of Gaussian processes: an
exploration-exploitation approach},
author = {Andreas Krause and Carlos Guestrin},
booktitle = {International Conference on Machine Learning (ICML)},
month = {June},
year = 2007,
address = {Corvallis, Oregon},
wwwfilebase = {icml2007-krause-guestrin},
wwwtopic = {Sensor Networks}
}



full list