Neural Architecture Search based on Cartesian Genetic Programming Coding Method

TL;DR

This paper introduces CGPNAS, an evolutionary neural architecture search method based on Cartesian genetic programming, demonstrating competitive performance and good transferability for sentence classification tasks.

Contribution

It proposes a novel NAS approach using CGP with evolutionary strategies, highlighting the importance of attention functions and linear transformations.

Findings

Evolved architectures perform comparably to human-designed models.

Transferability experiments show accuracy drops below 5%.

Ablation study identifies key functions like Attention.

Abstract

Neural architecture search (NAS) is a hot topic in the field of automated machine learning and outperforms humans in designing neural architectures on quite a few machine learning tasks. Motivated by the natural representation form of neural networks by the Cartesian genetic programming (CGP), we propose an evolutionary approach of NAS based on CGP, called CGPNAS, to solve sentence classification task. To evolve the architectures under the framework of CGP, the operations such as convolution are identified as the types of function nodes of CGP, and the evolutionary operations are designed based on Evolutionary Strategy. The experimental results show that the searched architectures are comparable with the performance of human-designed architectures. We verify the ability of domain transfer of our evolved architectures. The transfer experimental results show that the accuracy deterioration is lower than 2-5%. Finally, the ablation study identifies the Attention function as the single key function node and the linear transformations along could keep the accuracy similar with the full evolved architectures, which is worthy of investigation in the future.