LifeScreen Analytics

Problem

Excessive screen time is difficult to quantify in terms of behavioral risk. This project focuses on analyzing usage patterns and predicting potential digital addiction risk.

Approach

• Performed EDA on screen time datasets to identify behavioral patterns
• Engineered features like session frequency and usage duration
• Trained ML models to classify addiction risk levels

Results

The model was able to identify high-risk usage patterns and provide interpretable insights for behavior monitoring.

What I learned

• Feature engineering plays a major role in behavioral ML problems
• EDA is critical before applying models
• Simple models can perform well with meaningful features

View Code