A Genetic Programming Approach to Designing Convolutional Neural Network Architectures

TL;DR

This paper introduces a genetic programming method to automatically design CNN architectures for image classification, reducing the need for expert knowledge and trial-and-error in model development.

Contribution

It presents a novel approach using Cartesian genetic programming with functional modules to optimize CNN structures for improved accuracy.

Findings

Automatically designed CNNs achieve competitive accuracy on CIFAR-10.

The method reduces manual effort in CNN architecture design.

Optimized architectures outperform some existing models.

Abstract

The convolutional neural network (CNN), which is one of the deep learning models, has seen much success in a variety of computer vision tasks. However, designing CNN architectures still requires expert knowledge and a lot of trial and error. In this paper, we attempt to automatically construct CNN architectures for an image classification task based on Cartesian genetic programming (CGP). In our method, we adopt highly functional modules, such as convolutional blocks and tensor concatenation, as the node functions in CGP. The CNN structure and connectivity represented by the CGP encoding method are optimized to maximize the validation accuracy. To evaluate the proposed method, we constructed a CNN architecture for the image classification task with the CIFAR-10 dataset. The experimental result shows that the proposed method can be used to automatically find the competitive CNN architecture compared with state-of-the-art models.