Student Social Media & Mental Health Analysis
This end-to-end data science project explores the impact of screen time, digital unlocks, study habits, and sleep on student mental health. To demonstrate core math proficiency, the machine learning models—OLS Linear Regression (predicting Mental Health Score) and a Gini Impurity Decision Tree Classifier (predicting Stress Risk)—were implemented from scratch in Python and deployed directly in this interactive dashboard.
Most Used Social Platforms
Distribution of dominant platforms among students
Average Screen Time by Stress Level
Showing direct link between daily hours and stress severity
Screen Time vs. Mental Health Score
Scatter distribution of selected samples with stress-level highlights
Feature Correlation Matrix
Pearson correlation coefficients among key metrics
Predictive Analysis & Risk Assessment Sandbox
Input personalized metrics to evaluate mental health score and stress risk using the custom-trained models.
Risk & Score Output
Live Inference ActiveCohorts Dataset Explorer
Inspecting sample student logs (Loaded directly from raw CSV data)
| Age | Gender | Platform | Usage (Hrs) | Daily Unlocks | Study Hours | Sleep (Hrs) | Stress Level | MH Score |
|---|