Gender Classification with A NovelConvolutional Neural Network (CNN) Modeland Comparison with other MachineLearning and Deep Learning CNN Models
Yazarlar (3)
Doç. Dr. Özkan İNİK
Tokat Gaziosmanpaşa Üniversitesi, Türkiye
Konya Teknik Üniversitesi, Türkiye
Selçuk Üniversitesi, Türkiye
| Makale Türü | Özgün Makale |
| Makale Alt Türü | Diğer hakemli uluslarası dergilerde yayınlanan tam makale |
| Dergi Adı | Journal of Industrial Engineering Research |
| Dergi ISSN | 2077-4559 |
| Makale Dili | İngilizce |
| Basım Tarihi | 12-2018 |
| Cilt No | 4 |
| Sayı | 4 |
| Sayfalar | 57 / 63 |
| Makale Linki | http://www.iwnest.com/old/JIER/2018/December/57-63.pdf |
| Özet |
| Processing and analyzing of the big data has become difficult with the increasing data size in recent years. In 2006, Geoffrey Hinton, one of the machine learning pioneers, represented a deep learning model at the University of Toronto. The basis of the deep learning model is based on the use of large number of hidden layers in artificial neural networks. In previous years, increasing the number of hidden layers was not preferred because of the complexity in calculation process. With graphical processing unit (GPU) technology, calculations are done more quickly, so the popularity of deep learning has begun to increase again. Convolutional Neural Network (CNN) is a kind of deep learning model represented for the large-scale image classification. CNN is the most widely used deep learning model in feature learning, recognition, and classification. The purpose of this study is to classify gender. For this aim, a new CNN model has been designed. The Adience data set was used for training and testing of this designed CNN model. As a result of the experimental studies, gender classification was done with 88.5% accuracy rate. It has been seen that the best accuracy value is obtained by the proposed model when compared with machine learning method and other CNN model. |
| Anahtar Kelimeler |
