This study aims to identify the factors that predict academic performance in mathematics, science, and reading subjects among Turkish secondary school students. Using data from the OECD's PISA 2018 survey, including student- and school-level variables as well as PISA 2018 test scores, we employed a range of supervised regression-based machine learning methods to compare their predictive performance. Our results demonstrate that the boosting regression tree (BRT) method outperforms other methods, including multiple linear regression, ridge regression, LASSO, elastic net regression, bagging, and random forest regression trees. BRT highlights the importance of general secondary education programs over vocational and technical education in predicting academic achievement. Moreover, both student-level and school-level characteristics are shown to be significant predictors of academic performance in all subject areas. These findings contribute to the development of evidence-based educational policies in Turkey.