Achieving Deep Continual Learning via Evolution

TL;DR

This paper introduces Evolving Continual Learning (ECL), a novel framework inspired by collective learning, which maintains and evolves a diverse population of neural networks to improve continual learning performance.

Contribution

ECL is the first to use a population of evolving neural networks to address continual learning, combining architecture search and specialization for each task.

Findings

ECL outperforms state-of-the-art CL methods in experiments.

ECL effectively maintains stability and plasticity through expert isolation and evolution.

The approach demonstrates significant improvements in continual learning benchmarks.

Abstract

Deep neural networks, despite their remarkable success, remain fundamentally limited in their ability to perform Continual Learning (CL). While most current methods aim to enhance the capabilities of a single model, Inspired by the collective learning mechanisms of human populations, we introduce Evolving Continual Learning (ECL), a framework that maintains and evolves a diverse population of neural network models. ECL continually searches for an optimal architecture for each introduced incremental task. This tailored model is trained on the corresponding task and archived as a specialized expert, contributing to a growing collection of skills. This approach inherently resolves the core CL challenges: stability is achieved through the isolation of expert models, while plasticity is greatly enhanced by evolving unique, task-specific architectures. Experimental results demonstrate that ECL significantly outperforms state-of-the-art individual-level CL methods. By shifting the focus from individual adaptation to collective evolution, ECL presents a novel path toward AI systems capable of CL.