Multilayer perceptron model for prostate cancer prediction

Abstract

We present a regression model based on multilayer perceptron (MLP) network for the level of prostate specific antigen. Finally, the results of the MLP-network are compared with other regression models; the conclusion is that the MLP-model is approximately 10 times more accurate that the other models.

Keywords:

• Neural network
• Regression model
• Prostate specific antigen