Mind Waves and Machines: Predicting Technologial Addiction with Electroencephalography

This study investigates the predictability of technological addiction using Electroencephalography data and machine learning models. Electroencephalography recordings from 1,000 participants were classified into three groups: addicted, not addicted, and recovered from addiction through meditation. A range of predictive modeling techniques, including Random Forest, XGBoost, and K-Nearest Neighbor (KNN), were used to classify addiction stages.. K-Nearest Neighbor achieved the highest accuracy of 95.62%, demonstrating that Electroencephalography signals can differentiate between addiction and recovery states, supporting the role of meditation in recovery.