Skip to main content

Research Repository

Advanced Search

On Optimizing the Structure of Neural Networks Through a Compact Codification of Their Architecture

Lupión, Marcos; Cruz, N. C.; Paechter, B.; Ortigosa, P. M.

Authors

Marcos Lupión

N. C. Cruz

P. M. Ortigosa



Abstract

Neural networks stand out in Artificial Intelligence for their capacity of being applied to multiple challenging tasks such as image classification. However, designing a neural network to address a particular problem is also a demanding task that requires expertise and time-consuming trial-and-error stages. The design of methods to automate the designing of neural networks define a research field that generally relies on different optimization algorithms, such as population meta-heuristics. This work studies utilizing Teaching-Learning-based Optimization (TLBO), which had not been used before for this purpose up to the authors’ knowledge. It is widespread and does not have specific parameters. Besides, it would be compatible with deep neural network design, i.e., architectures with many layers, due to its conception as a large-scale optimizer. A new encoding scheme has been proposed to make this continuous optimizer compatible with neural network design. This method, which is of general application, i.e., not linked to TLBO, can represent different network architectures with a plain vector of real values. A compatible objective function that links the optimizer and the representation of solutions has also been developed. The performance of this framework has been studied by addressing the design of an image classification neural network based on the CIFAR-10 dataset. The achieved result outperforms the initial solutions designed by humans after letting them evolve.

Presentation Conference Type Conference Paper (Published)
Conference Name Metaheuristics: 14th International Conference, MIC 2022
Start Date Jul 11, 2022
End Date Jul 14, 2022
Acceptance Date Jun 1, 2022
Online Publication Date Feb 23, 2023
Publication Date 2023-02
Deposit Date Jun 12, 2023
Publisher Springer
Pages 133-142
Series Title Lecture Notes in Computer Science
Series Number 13838
Series ISSN 1611-3349
Book Title Metaheuristics: 14th International Conference, MIC 2022, Syracuse, Italy, July 11–14, 2022, Proceedings
ISBN 978-3-031-26503-7
DOI https://doi.org/10.1007/978-3-031-26504-4_10
Keywords Artificial intelligence, Neural network architecture optimization, Meta-heuristics, Teaching-Learning-based optimization