Impacts of Darwinian Evolution on Pre-trained Deep Neural Networks

TL;DR

This paper introduces an evolutionary-inspired framework for optimizing deep neural networks, demonstrating reduced overfitting, lower time complexity, and improved performance on large datasets compared to traditional back-propagation methods.

Contribution

It proposes a novel evolutionary-based optimization method for deep neural networks, bridging biological evolution concepts with neural network training.

Findings

Reduces overfitting in neural networks.

Lower time complexity than back-propagation.

Effective on large datasets and deep models.

Abstract

Darwinian evolution of the biological brain is documented through multiple lines of evidence, although the modes of evolutionary changes remain unclear. Drawing inspiration from the evolved neural systems (e.g., visual cortex), deep learning models have demonstrated superior performance in visual tasks, among others. While the success of training deep neural networks has been relying on back-propagation (BP) and its variants to learn representations from data, BP does not incorporate the evolutionary processes that govern biological neural systems. This work proposes a neural network optimization framework based on evolutionary theory. Specifically, BP-trained deep neural networks for visual recognition tasks obtained from the ending epochs are considered the primordial ancestors (initial population). Subsequently, the population evolved with differential evolution. Extensive experiments are carried out to examine the relationships between Darwinian evolution and neural network optimization, including the correspondence between datasets, environment, models, and living species. The empirical results show that the proposed framework has positive impacts on the network, with reduced over-fitting and an order of magnitude lower time complexity compared to BP. Moreover, the experiments show that the proposed framework performs well on deep neural networks and big datasets.