New M.E.Thesis Submitted from IT

Intelligent multi-layer perceptron algorithm for disease classification application, By Gursharan Singh, IT

Abstract

Decision tree based techniques are being continuously evolved for classification of medical datasets. Different algorithms based on soft computing and hard computing has been developed to apply on medical datasets. Hybrid techniques find their way out to search for the solution of cancer databases. Some are inspired by nature and some are inspired by biological phenomenon. Biological inspired methods are successfully developed and applied for different medical problems. In current research, an intelligent technique inspired from neural network, supervised multilayer perceptron decision tree have been applied and evaluated successfully to classify lung cancer based datasets. A comparison has been made with other techniques to check the effectiveness of the proposed method. Simulation results shows that proposed technique achieved appropriate accuracy to classify cancer data sets which is more as compared to other techniques. TP rate, ROC and Precision is highest for proposed method amongst other method. Hence, proposed algorithm is optimum to classify cancer related datasets.