Automatic detection of erythemato-squamous diseases using K- Nearest Neighbor Algorithm

Data mining in healthcare is an emerging field of high importance for providing prognosis and a deeper understanding of medical data. Healthcare data mining attempts to solve real world health problems in diagnosis and treatment of diseases. One of the most important usage of data mining in machine learning domain is differential diagnosis. The differential diagnosis of erythemato-squamous diseases is a difficult problem in dermatology. The goal of this research is gaining to a high performance differential diagnosing of erythemato-squamous disease by using KNN Learning algorithms. UCI data base including 366 records has been used in this research. KNN algorithm was implemented and accuracy of system was compared with other methods used to diagnose such disease previously the highest accuracy was related to k=5. Sensitivity, specificity and accuracy in the best condition were respectively obtained 1, 1, and 0.98 and compared with other methods. KNN approach with preprocessing based on Relief-F strategy and Cross Fold modeling could be suggested as an effective data mining method to classify and diagnosis different diseases.