A genetic programming approach to designing convolutional neural network architectures

Masanori Suganuma, Shinichi Shirakawa, Tomoharu Nagao

Research output: Chapter in Book/Report/Conference proceedingConference contribution

209 Citations (Scopus)

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 con-volutional 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.

Original languageEnglish
Title of host publicationGECCO 2017 - Proceedings of the 2017 Genetic and Evolutionary Computation Conference
PublisherAssociation for Computing Machinery, Inc
Pages497-504
Number of pages8
ISBN (Electronic)9781450349208
DOIs
Publication statusPublished - 2017 Jul 1
Externally publishedYes
Event2017 Genetic and Evolutionary Computation Conference, GECCO 2017 - Berlin, Germany
Duration: 2017 Jul 152017 Jul 19

Publication series

NameGECCO 2017 - Proceedings of the 2017 Genetic and Evolutionary Computation Conference

Other

Other2017 Genetic and Evolutionary Computation Conference, GECCO 2017
CountryGermany
CityBerlin
Period17/7/1517/7/19

Keywords

  • Convolutional neural network
  • Deep learning
  • Designing neural network architectures
  • Genetic programming

ASJC Scopus subject areas

  • Software
  • Computer Science Applications
  • Computational Theory and Mathematics

Fingerprint Dive into the research topics of 'A genetic programming approach to designing convolutional neural network architectures'. Together they form a unique fingerprint.

Cite this