An Improve KNN Method for Classification of Sexually Transmitted Diseases

This Sexually Transmitted Diseases (STD) are diseases caused by infections that are transmitted through sexual contact. One way to overcome the problem of sexually transmitted diseases is to create an application that can help people in early detection. One method that can analyze PMS based on its symptoms is the K-NN method. However, the K-NN method still produces 90% accuracy using venereal disease datasets. Several ways to improve the accuracy of this method include using different distance calculations and using weights as calculations, one of which is by improving KNN. In this study, the WBCKNN modification was carried out on the same dataset, namely the venereal disease dataset. Modification of this method is done by adding leave-one-out cross-validation (LOOCV), without LOOCV, undersampling and oversampling. The proposed algorithm is evaluated and compared with the KNN, WKNN, BCKNN methods that have been analyzed previously. This study shows that the WBCKNN method is effective with an accuracy of 94.24% without LOOVC, the addition of LOOCV produces an accuracy of 90.70%, with the data type undersampling without LOOVC produces an accuracy of 99.90% and oversampling without LOOVC produces an accuracy of 98.47%. In this study the WBCKNN method produced high accuracy compared to the previous method.