Symbolic-Regression Boosting

TL;DR

SyRBo enhances symbolic regression by integrating a small number of boosting stages, leading to significant performance improvements with minimal additional computational cost, and can be easily implemented on existing symbolic regressors.

Contribution

Introduces SyRBo, a novel boosting method that replaces weak learners with stronger symbolic regressors, improving accuracy with minimal complexity.

Findings

Statistically significant improvements over 98 datasets

Boosting with 2-5 stages yields better results

Easy to implement as an add-on to existing regressors

Abstract

Modifying standard gradient boosting by replacing the embedded weak learner in favor of a strong(er) one, we present SyRBo: Symbolic-Regression Boosting. Experiments over 98 regression datasets show that by adding a small number of boosting stages -- between 2--5 -- to a symbolic regressor, statistically significant improvements can often be attained. We note that coding SyRBo on top of any symbolic regressor is straightforward, and the added cost is simply a few more evolutionary rounds. SyRBo is essentially a simple add-on that can be readily added to an extant symbolic regressor, often with beneficial results.