Adequate assistance in the learning process is important using accurate estimation of students' academic performance based on new emerging techniques, and discover new knowledge, finding meaningful variables answers educational problems. Data contains hidden information for analyzing, extracting information and knowledge to find patterns, and using this for decision-making. WEKA (Waikato Environment for Knowledge Analysis) is used and Na?ve Bayesian Classification Method is implemented, as the pre-processing mechanism for 10-fold cross-validation where classification models were generated, cross-validation method, and percentage split were used to evaluate the efficiency of this algorithm. Linear Regression is used to identify the significant predictors that affect the students' academic performance. The simulation results show that with Na?ve Bayes Classification Method algorithm for classification with a correctly classified instance of 93.48% and incorrectly classified instance of 6.52%, this result indicates that with the classification method the accuracy level of prediction increases. To validate the generated model, the experiments were conducted using real data. The result is a good model intended to be used in the school decision support system in predicting learners' academic performance.