Comparative Analysis of Logistic Regression and Gaussian Naïve Bayes Classifiers in Binary Classification Tasks using MATLAB Pseudo-code

This research presents a comparative evaluation of two prominent supervised learning algorithms, Logistic Regression (LR) and Gaussian Naïve Bayes (GNB), on binary classification tasks. Both algorithms are widely used due to their interpretability and efficiency. Using the Pima Indians Diabetes Dataset from the UCI Machine Learning Repository, the study assesses and compares their classification performance based on accuracy, precision, recall, F1-score, and ROC-AUC. Results show that Logistic Regression demonstrates superior performance on linearly separable data, while Gaussian Naïve Bayes exhibits greater robustness to data distributional assumptions. These findings contribute to understanding model suitability for specific binary classification contexts.