CART MAKİNA ÖĞRENME ALGORİTMASINDA İYİLEŞTİRME VE BANKNOT DENETLEME VERİSİNDE UYGULAMA

Abstract

In this scientific study, it is aimed to examine all processes while performing classification process with CART algorithm and to make algorithmic improvements by using float type data obtained from Kaggle platform. In cross validation phase, it is expected that the tree structure will be educated more accurately by selecting the training data more accurately. When testing and training data is subdivided, performing these procedures according to certain criteria will increase the stability of the system and increase the success rate, especially during the training phase. Algorithmic improvements will be made in order to increase the effect of the data to be used in the division of data set n sub-section on the classification result.