Diagnosis of disease by using data mining methods on real medical data

Nowadays data mining plays a significant role in solving most of the critical problems. Especially in medical field, storage of medical data in databases creates a large mass of data which is being the most preferred application area for data mining methods. Obtaining information quickly and efficiently through the biomedical systems which use data mining techniques, provide a great benefit to physicians and patients.In this thesis k-nearest neighbor (KNN) and k-means algorithms have been investigated in detail. Also using these algorithms an assistant decision-making system which is available to physicians in the medical field, can predict for the diagnosis of dermatological diseases and provide an opportunity to analyze the relationships between characteristics of patient records has been designed and carried out.Keywords Data mining, K nearest neighbour, K-means, Biomedicine, Medical Informatics, Dermatology.