DEA-neural network approach to solve binary classification problems

Subject Areas : International Journal of Data Envelopment Analysis

Saeeid Kashanifar ¹ , Mona Farahnak Roudsary ²

1 - Department of Industrial Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran
2 - Department of Industrial Engineering, Gazvin Branch, Islamic Azad University, Gazvin, Iran

Received: 2022-04-12 Accepted : 2022-10-24 Published : 2022-11-01

Keywords: Binary classification, Data Envelopment Analysis, Radial basis function, Linear programming problem,

Abstract :

In this paper, we propose a new hybrid neural network including Data Envelopment Analysis(DEA) and Radial Basis Function Network (RBFN)for binary classification problems. In the supervised learning phase of the neural network, the additive model is used to learn the classification function and Gaussian Radial Basis Function (GRBF) is used in the unsupervised learning phase of the neural network. Compared with the existing RBFN-DEA model for solving classification problems, the proposed model has low CPU time and can be applied to solve classification problems with negative data.