<![CDATA[Makassar City has a varied monsoon rainfall pattern, so rainfall prediction is an important challenge in disaster mitigation and resource management. Data mining techniques such as classification with the Classification and Regression Trees (CART) algorithm can be used to classify rainfall and analyze historical data, but the risk of overfitting high-dimensional data requires dimension reduction such as Principal Component Analysis (PCA). To improve accuracy, the Output Statistics Model (MOS) approach that combines numerical data and observations is also used. The results of dimension reduction using the Principal Component Analysis (PCA) method showed that of the initial seven variables, only three main components (, , and ) were retained because they had eigenvalues greater than 1 and were able to explain the data variance significantly. The decision tree model that was formed resulted in an accuracy rate of 72.34% in training data. Where the model can classify most of the training data into the correct rainfall category. In the data testing, the model was able to achieve an accuracy level of 71.43%, which shows that the model has good generalization ability to new data and does not experience overfitting.]]>
Copyrights © 2026
