OPTIMIZATION OF RBF-SVM HYPERPARAMETERS USING GENETIC ALGORITHM FOR FACE RECOGNITION

Authors

  • Y Ibrahim DEPT. OF COMPUTER ENGINEERING, AHMADU BELLO UNIVERSITY, ZARIA, KADUNA STATE, NIGERIA
  • E Okafor DEPT. OF COMPUTER ENGINEERING, AHMADU BELLO UNIVERSITY, ZARIA, KADUNA STATE, NIGERIA
  • B Yahaya DEPT. OF COMPUTER ENGINEERING, AHMADU BELLO UNIVERSITY, ZARIA, KADUNA STATE, NIGERIA

Keywords:

Face Recognition, Feature Extraction, Local Binary Patterns, Transfer Learning, Genetic Algorithm and Support Vector Machines.

Abstract

Manual grid-search tuning of machine learning hyperparameters is very time-consuming. Hence, to curb this problem, we propose the use of a genetic algorithm (GA) for the selection of optimal radial-basis-function based support vector machine (RBF-SVM) hyperparameters; regularization parameter C and cost-factor ?. The resulting optimal parameters were used during the training of face recognition models. To train the models, we independently extracted features from the ORL face image dataset using local binary patterns (handcrafted) and deep learning architectures (pretrained variants of VGGNet). The resulting features were passed as input to either linear-SVM or optimized RBF-SVM. The results show that the models from optimized RBF-SVM combined with deep learning or hand-crafted features yielded performances that surpass models obtained from Linear-SVM combined with the aforementioned features in most of the data splits. The study demonstrated that it is profitable to optimize the hyperparameters of an SVM to obtain the best classification performance.

 

http://dx.doi.org/10.4314/njt.v39i4.27

Downloads

Published

2020-09-30

Issue

Vol. 39 No. 4 (2020): NIGERIAN JOURNAL OF TECHNOLOGY, VOL. 39, NO. 4, OCTOBER 2020

Section

Computer, Telecommunications, Software, Electrical & Electronics Engineering

License

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

The contents of the articles are the sole opinion of the author(s) and not of NIJOTECH.

NIJOTECH allows open access for distribution of the published articles in any media so long as whole (not part) of articles are distributed.

A copyright and statement of originality documents will need to be filled out clearly and signed prior to publication of an accepted article. The Copyright form can be downloaded from http://nijotech.com/downloads/COPYRIGHT%20FORM.pdf while the Statement of Originality is in http://nijotech.com/downloads/Statement%20of%20Originality.pdf

 

For articles that were developed from funded research, a clear acknowledgement of such support should be mentioned in the article with relevant references. Authors are expected to provide complete information on the sponsorship and intellectual property rights of the article together with all exceptions.

 

It is forbidden to publish the same research report in more than one journal.

How to Cite

OPTIMIZATION OF RBF-SVM HYPERPARAMETERS USING GENETIC ALGORITHM FOR FACE RECOGNITION. (2020). Nigerian Journal of Technology, 39(4), 1190-1197. https://nijotech.com/index.php/nijotech/article/view/2460