A Novel Neighborhood Rough Set Based Classification Approach for Medical Diagnosis

Medical datasets consume enormous amount of information about the patients, diseases and the physicians. Diseases diagnosis required many expensive tests to predict the diseases. Cost of disease prediction and diagnosis can be reduced by applying machine learning and data mining methods. Disease prediction and decision making plays asignificant role in medical diagnosis. In this study, a novel neighborhood rough set classification approach is presented to deal with medical datasets. Five benchmarked medical datasets have been used in this research work for studying the impact of proposed work in decision making.Experimental resultof the proposed classification algorithm is compared with other existing approaches such as rough set, Kth–nearest neighbor, support vector machine, Back propagation algorithm and multilayer perceptron to conclude that the proposed approach is cheaper way for disease prediction and decision making. The performance of classification algorithms measured based on various classification accuracy measures.