We will compare the best results of the raw data with CHI and PCA.Compared to the raw data, CHI and CHI-PCA improved in the computations of LOG, MPC, NB. However, the performance decreased with DT, and GBT.CHI-PCA-NB presented the worst value, with an accuracy of 68.4%, an F1 score of 75.7%, an MCC of 37.7%, and a Kappa of 33.7%.Cleveland dataset obtained the best results using CHI-PCA (Fig.GBT presented a pattern when applied with CHI and PCA as shown in Fig.2 and Fig. 3).2.