Authors:
Olivier Teytaud
1
;
Sylvain Gelly
1
and
Jérémie Mary
2
Affiliations:
1
TAO (Inria, Univ. Paris-Sud, UMR CNRS-8623), France
;
2
TAO (Inria, Univ. Paris-Sud, UMR CNRS-8623); Grappa (Inria Univ. Lille), France
Keyword(s):
Intelligent Control Systems and Optimization, Machine learning in control applications, Active learning.
Related
Ontology
Subjects/Areas/Topics:
Informatics in Control, Automation and Robotics
;
Intelligent Control Systems and Optimization
;
Machine Learning in Control Applications
Abstract:
We study active learning as a derandomized form of sampling. We show that full derandomization is not suitable in a robust framework, propose partially derandomized samplings, and develop new active learning methods (i) in which expert knowledge is easy to integrate (ii) with a parameter for the exploration/exploitation dilemma (iii) less randomized than the full-random sampling (yet also not deterministic). Experiments are performed in the case of regression for value-function learning on a continuous domain. Our main results are (i) efficient partially derandomized point sets (ii) moderate-derandomization theorems (iii) experimental evidence of the importance of the frontier (iv) a new regression-specific user-friendly sampling tool less-robust than blind samplers but that sometimes works very efficiently in large dimensions. All experiments can be reproduced by downloading the source code and running the provided command line.