Entropy from Machine Learning
Abstract
We translate the problem of calculating the entropy of a set of binary configurations/signals into a sequence of supervised classification tasks. Subsequently, one can use virtually any machine learning classification algorithm for computing entropy. This procedure can be used to compute entropy, and consequently the free energy directly from a set of Monte Carlo configurations at a given temperature. As a test of the proposed method, using an off-the-shelf machine learning classifier we reproduce the entropy and free energy of the 2D Ising model from Monte Carlo configurations at various temperatures throughout its phase diagram. Other potential applications include computing the entropy of spiking neurons or any other multidimensional binary signals.
Cite
@article{arxiv.1909.10831,
title = {Entropy from Machine Learning},
author = {Romuald A. Janik},
journal= {arXiv preprint arXiv:1909.10831},
year = {2019}
}
Comments
10 pages, 2 figures; v2: reference added, minor notational improvement; v3: reference added, general comments in section 3