Exploration and Exploitation in Symbolic Regression using Quality-Diversity and Evolutionary Strategies Algorithms
Neural and Evolutionary Computing
2019-06-11 v1 Machine Learning
Abstract
By combining Genetic Programming, MAP-Elites and Covariance Matrix Adaptation Evolution Strategy, we demonstrate very high success rates in Symbolic Regression problems. MAP-Elites is used to improve exploration while preserving diversity and avoiding premature convergence and bloat. Then, a Covariance Matrix Adaptation-Evolution Strategy is used to evaluate free scalars through a non-gradient-based black-box optimizer. Although this evaluation approach is not computationally scalable to high dimensional problems, our algorithm is able to find exactly most of the targets extracted from the literature on which we evaluate it.
Cite
@article{arxiv.1906.03959,
title = {Exploration and Exploitation in Symbolic Regression using Quality-Diversity and Evolutionary Strategies Algorithms},
author = {J. -P. Bruneton and L. Cazenille and A. Douin and V. Reverdy},
journal= {arXiv preprint arXiv:1906.03959},
year = {2019}
}
Comments
11 pages, 7 figures, 5 tables