MaxEnt Queries and Sequential Sampling
Disordered Systems and Neural Networks
2009-10-31 v1 Statistical Mechanics
Abstract
In this paper we pose the question: After gathering N data points, at what value of the control parameter should the next measurement be done? We propose an on-line algorithm which samples optimally by maximizing the gain in information on the parameters to be measured. We show analytically that the information gain is maximum for those potential measurements whose outcome is most unpredictable, i.e. for which the predictive distribution has maximum entropy. The resulting algorithm is applied to exponential analysis.
Keywords
Cite
@article{arxiv.cond-mat/0010104,
title = {MaxEnt Queries and Sequential Sampling},
author = {Peter Riegler and Nestor Caticha},
journal= {arXiv preprint arXiv:cond-mat/0010104},
year = {2009}
}
Comments
Presented at MaxEnt 2000, the 20th International Workshop on Bayesian Inference and Maximum Entropy Methods (July 8-13, 2000, Gif-sur-Yvette, France) To be published in the Proceedings, Ali Mohammad-Djafari Ed