Thalassemic patient classification using a neural network and genetic programming

This paper presents the use of a genetic programming (GP) system called STROGANOFF and a multilayer perceptron for thalassemic patient classification. The interested problem covers the test samples from normal subjects and that from different types of thalassemic patient and thalassemic trait. The features, which are the characteristics of red blood cell, reticulocyte and blood platelet extracted from the blood samples, are used as input to the classifiers. The results indicate that the performance of the GP-generated classification trees is approximately equal to that of the multilayer perceptrons with one hidden layer. In contrast, the multilayer perceptrons with two hidden layers outperform GP-generated classification trees. Nonetheless, the structure of the classification trees reveals that the characteristics of blood platelet have no effects on the classification performance. This helps to reduce the required input features for the task and make further improvements possible.