Published Paper Details:

This paper presents a comparative study of advanced machine learning algorithms for prime number classification across diverse numerical ranges. Seven models Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), Support Vector Machine (SVM), k-Nearest Neighbors (kNN), XGBoost (XGB), and LightGBM (LGBM) are systematically evaluated in terms of both predictive accuracy and computational efficiency across eight distinct prime number intervals. The assessment incorporates key performance metrics such as Accuracy, Precision, Recall, F1-Score, ROC-AUC, average training time per epoch, and the average number of epochs required for convergence. Findings from the analysis emphasize the inherent trade-offs between accuracy and efficiency, providing practical guidance for selecting optimal models in large-scale classification scenarios. The results reveal noticeable variations in model performance based on both the prime number range and the chosen algorithm. Ensemble models like XGBoost and LightGBM consistently achieved higher accuracy in larger ranges, whereas Logistic Regression and Decision Trees showed limited scalability. k-Nearest Neighbors performed well in smaller ranges but lost efficiency as the dataset size increased. Overall, the study underscores the balance between accuracy and computational cost, providing useful guidance for selecting suitable machine learning models in prime number classification and broader computational number theory tasks.