Development of an Efficient Classifier for Classification of Liver Patient with Feature Selection

Diagnosis of health conditions is a very challenging task in field of medical science. In medical science, day by day data is increasing continuously and creates problem to identify the accurate diseases. Data mining based classification plays very important role in classification of data. In this research work we have used various data mining based classification technique to develop the classifier for classification of liver and non liver patient. We have used techniques like C4.5, Random Forest (RF), Multilayer Perceptron(MLP), Classification and Regression Technique (CART) and applied all these techniques on liver patient data collected from UCI repository.In this paper we have used ensemble model to develop the robust classification model which gives higher classification accuracy compare to its individual model. We have also used Information gain feature selection technique is applied on best model ensemble of C4.5, Random Forest and CART which gives 76.02% of accuracy with 3 numbers of features.