@article{article_1761875, title={Teen Smartphone Addiction Level Prediction Using Machine Learning Models: A Comparative Study}, journal={Osmaniye Korkut Ata Üniversitesi Fen Bilimleri Enstitüsü Dergisi}, volume={9}, pages={2018–2047}, year={2026}, DOI={10.47495/okufbed.1761875}, url={https://izlik.org/JA92YZ62NU}, author={Doğan, Alican}, keywords={Akıllı telefon bağımlılığı, Makine öğrenimi, Sınıflandırma, Sinir ağları}, abstract={Smartphone addiction among teenagers has become a growing concern due to its effect on brain status, school grades, and daily routines. This study investigates the prediction of smartphone addiction severity using a behavioral dataset that includes features such as daily time spent on social media, gaming, education, physical activity, academic metrics, and cognitive indicators. The target variable, Addiction_Level, was modeled as a classification problem after appropriate binning. Eight different machine learning models were implemented and compared, including Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Gradient Boosting (GB), XGBoost, and Multilayer Perceptron (MLP). Among these, LR and MLP methods obtained the highest accuracy, with 98.67% and 97.17%, respectively, outperforming more complex ensemble methods. The results suggest that simple yet robust models can effectively predict smartphone addiction levels based on behavioral patterns and academic indicators. This study offers valuable insights for educators, psychologists, and policy-makers aiming to detect and mitigate smartphone overuse in adolescent populations through data-driven strategies.}, number={4}