Deep Learning-Based Operators for Evolutionary Algorithms

TL;DR

This paper introduces two deep learning-based genetic operators, a crossover and a mutation, that improve evolutionary algorithms by leveraging neural networks for gene selection and mutation.

Contribution

The paper proposes novel, domain-independent deep learning operators for genetic algorithms and programming, enhancing evolutionary search capabilities.

Findings

Deep Neural Crossover effectively combines genes using reinforcement learning.

BERT mutation improves individual fitness by strategic node replacement.

Both operators outperform traditional methods in experiments.

Abstract

We present two novel domain-independent genetic operators that harness the capabilities of deep learning: a crossover operator for genetic algorithms and a mutation operator for genetic programming. Deep Neural Crossover leverages the capabilities of deep reinforcement learning and an encoder-decoder architecture to select offspring genes. BERT mutation masks multiple gp-tree nodes and then tries to replace these masks with nodes that will most likely improve the individual's fitness. We show the efficacy of both operators through experimentation.