Comparative analysis of the prediction and classification accuracy of artificial neural networks with respect to traditional statistical methods

  • Carlos Belman-López TecNM in Celaya
  • José Alfredo Jiménez García
  • José Antonio Vázquez López
Keywords: Artificial Neural Networks, statistic, multilayer perceptron, regression, discriminant analysis

Abstract

ANNs are flexible tools that have shown utility as function estimators and prediction methods in Data Mining, Machine Learning, among others. The aim of this research is to carry out comparative analysis of the prediction and classification accuracy between artificial neural networks and traditional statistical methods. To carry out this comparative analysis, different cases of prediction and supervised classification were selected. The samples of the different cases were divided into samples for training and validation. Finally, a comparative analysis of the prediction and classification accuracy between artificial neural networks and traditional statistical methods was obtained, using the sum of squared errors (SSE), the coefficient of determination (R2), among others.

Published
2021-04-15
How to Cite
Belman-López, C., Jiménez García, J. A., & Vázquez López, J. A. (2021). Comparative analysis of the prediction and classification accuracy of artificial neural networks with respect to traditional statistical methods. International Journal of Combinatorial Optimization Problems and Informatics, 12(2), 7-15. Retrieved from https://ijcopi.org/ojs/article/view/195
Section
Articles