Predicting the Impact of Mobile Phone Usage on Academic Performance Using Machine Learning: A Real-Time Digital Wellness Approach

Mobile phone addiction, Academic performance, Machine learning, Digital wellness, CNN-LSTM, Attention mechanism, Educational data mining, Risk detection

Authors

  • S. Vimala PhD Scholar, Department of Computer Science, St.Joseph’s College(Autonomous), Tiruchirappalli –Affiliated to Bharathidasan University, Tamil Nadu, India.
  • Dr. G. Arockia Sahaya Sheela Assistant Professor, Department of Computer Science, St.Joseph’s College(Autonomous), Tiruchirappalli –Affiliated to Bharathidasan University, Tamil Nadu, India.
February 24, 2026
February 24, 2026

Downloads

Purpose: This study investigates the impact of mobile phone usage on academic performance among students using machine learning algorithms, with a focus on developing a predictive model for real-time digital wellness risk detection and educational interventions. Methods: A comprehensive dataset of 5,000 students was collected through surveys and academic records from 2023-2024. The study employed advanced machine learning techniques including Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM) networks with attention mechanisms, and traditional algorithms (Random Forest, Gradient Boosting, Naive Bayes) to predict academic outcomes based on mobile phone usage patterns. Feature engineering incorporated screen time, app usage categories, study-to-phone ratios, and sleep duration metrics.  Results: The CNN-LSTM model with attention mechanism achieved the highest predictive accuracy of 92%, significantly outperforming traditional models. Students exceeding 4 hours daily on non-educational mobile applications showed a 20% decrease in academic performance. The study-to-phone usage ratio emerged as the most significant predictor of academic outcomes, while educational app usage demonstrated marginal positive correlations with performance.  Machine learning algorithms effectively predict mobile phone usage impact on student performance, enabling early intervention strategies for digital wellness. The findings support the development of real-time risk detection systems for educational institutions to implement targeted interventions and promote balanced technology use among students.