Fitness Approximation through Machine Learning

TL;DR

This paper introduces a machine learning-based method for fitness approximation in genetic algorithms, dynamically adapting to the evolutionary process to improve runtime efficiency with minimal impact on solution quality.

Contribution

The paper proposes a novel, adaptable fitness approximation technique using ML models within GAs, enhancing efficiency while maintaining solution quality.

Findings

Significant reduction in evolutionary runtime.

Fitness scores comparable to full evaluations.

Effective in costly fitness computation domains.

Abstract

We present a novel approach to performing fitness approximation in genetic algorithms (GAs) using machine-learning (ML) models, through dynamic adaptation to the evolutionary state. Maintaining a dataset of sampled individuals along with their actual fitness scores, we continually update a fitness-approximation ML model throughout an evolutionary run. We compare different methods for: 1) switching between actual and approximate fitness, 2) sampling the population, and 3) weighting the samples. Experimental findings demonstrate significant improvement in evolutionary runtimes, with fitness scores that are either identical or slightly lower than that of the fully run GA -- depending on the ratio of approximate-to-actual-fitness computation. Although we focus on evolutionary agents in Gymnasium (game) simulators -- where fitness computation is costly -- our approach is generic and can be easily applied to many different domains.