Students’ academic performance evaluation is fundamental in education data analytics for insightful decision makings in tertiary institutions of high learning. This research endeavors to construct a robust predictive machine learning model for anticipating student academic performance in an Economics-based course, specifically Econometrics. Leveraging demographic features and academic performance data obtained from the Economics Department Examination Unit Records at Prince Abubakar Audu University Nigeria, this study delves into the application of machine learning classifiers, including Naïve Bayes, K-Nearest Neighbor, Random Forest, and Support Vector Machine. The academic dataset, comprising 897 samples and 13 attributes, underwent classification analysis using a rigorous 80%-20% percentage split partitioning on the Weikato Environment for Knowledge Analysis (WEKA) machine learning model processing platform for training and testing processes. Feature selection methods, namely the correlation coefficient and recursive features elimination, were deployed to identify an optimal combination of predictor variables. The variables such as Intermediate Economic Statistic, Introduction to Mathematics for Economists, Intermediate Mathematics for Economics, Principles and Theories of Economics and Macroeconomics Analysis were identified as crucial in determining students' performance in Econometrics. The classification analysis conducted on this optimal feature set revealed that the Support Vector Machine algorithm outperformed other classifiers, achieving an impressive accuracy of 85.48%. Consequently, this study advocates for personalize teaching of those identified optimal predictor variables to enhance students’ academic performance and reduce student’s failure rate in Econometrics.