Diagnosis of Breast Cancer by Combining the Perceptron Neural Network Data Mining Techniques and Artificial Immune System

Abstract: Breast cancer is one of the most common cancers with a high mortality rate among women, which is the most common cause of death in the female population. Early diagnosis of breast cancer increases patient's chance of survival from 56% to over 86%. In this paper we used from new method NNCAISi that combination of two perceptron neural network algorithm and artificial immune system algorithm for more accurate diagnosis of tumors in breast cancer. In the combined model artificial neural network weights are trained by artificial immune algorithm and the best antibodies in the memory cell are returned as the optimized weight to the neural network. The data sets that used is from the University of California UCI website and the results are evaluated with three criteria for accuracy, precision and recall and it is compared and evaluated with some other algorithms.

Keywords: breast cancer, data mining, perceptron neural network, artificial immune system.