ABSTRACT

Classification has been used for predicting medical diagnosis. Classification methods can handle both numerical and categorical attributes. Gini index uses the method which biases multivalued attributes. When the number of classes are large, and the biases are increased, the Gini-based decision tree method is modified to overcome the known problems, by normalizing the Gini indexes by taking into account information about the splitting status of all attributes. Instead of using the Gini index for attribute selection ratios of Gini indexes are used and their splitting values in order to reduce the biases. Experiments are done on heart diseases dataset and Report of experimental graph is shown by comparing between the modified method and other known classification algorithms ID3, c4.5, Generalized Gini Index classifies relevant parts into various groups