Info-Evo: Using Information Geometry to Guide Evolutionary Program Learning

TL;DR

This paper introduces Info-Evo, an optimization method that employs information geometry and natural gradient search to enhance evolutionary program learning by guiding the search process along optimal paths.

Contribution

It presents a novel integration of information geometry with evolutionary algorithms, specifically applying natural gradient guidance to improve automated program learning.

Findings

Effective guidance of evolutionary search using natural gradient

Improved performance in automated program learning tasks

Successful integration with MOSES framework

Abstract

A novel optimization strategy, Info-Evo, is described, in which natural gradient search using nonparametric Fisher information is used to provide ongoing guidance to an evolutionary learning algorithm, so that the evolutionary process preferentially moves in the directions identified as "shortest paths" according to the natural gradient. Some specifics regarding the application of this approach to automated program learning are reviewed, including a strategy for integrating Info-Evo into the MOSES program learning framework.