Genetic-neuro approach for disease classification.

Venkatesan, P.; Premalatha, V.

International Journal of Science and Technology; 2012; 2; 473-478.  

Abstract: In this study, we introduce classification using artificial neural network whose architecture is trained by back propagation algorithm and input attributes are selected by evolutionary based genetic algorithm. Feature selection plays vital role in the applications of machine learning algorithms. For feature selection many approaches are used and the evolutionary based algorithms prove to be one of the efficient methods. We have used a breast cancer dataset for the evaluation of the new approach. The experimental result shows that the genetic-neuro system classification performs better than the conventional neural network.

Keywords: Classification; Feature selection; Genetic algorithm; Neural Network

Back to List of publications / Home