Digi Well

Digital Wellbeing Calculator for healthy Wellbeing

Comment

Digital-wellbeing

Mental Health and Digital Behavior

The dataset is taken from kaggle from https://www.kaggle.com/datasets/atharvasoundankar/mental-health-and-digital-behavior-20202024/data

Problem Statement

The goal of this project is to predict a user's digital_wellbeing_score based on various features like:

  • screen time
  • social media usage
  • notifications received
  • sleep hours
  • mood and focus levels
  • anxiety level

This helps in understanding how digital behavior affects mental wellness and can guide digital wellbeing recommendations.

Project Workflow

1. Data Reading

  • The dataset (mental_health_digital_behavior_data.csv) is loaded using pandas.
  • It contains 500 rows of user behavior and mental health scores from 2020–2024.

2. Data Exploration

  • Use .info(), .describe() and .head() to understand data types, ranges, and structure.
  • Checked for null values or data inconsistencies.

3. Data Visualization

  • Box plots and scatter plots were used to inspect distributions and detect outliers.
  • seaborn and matplotlib were used for visualizations.

4. Correlation Analysis

  • Used .corr() to analyze correlation of features with digital_wellbeing_score.
  • Found that anxiety_level was strongly negatively correlated.

5. Feature Selection

  • Selected all features except the target (digital_wellbeing_score) for training.

6. Data Preprocessing

  • No missing values, so minimal cleaning required.
  • All features were numerical, so no encoding was needed.

7. Model Training

Trained multiple regression models:

  • Linear Regression
  • Random Forest Regressor
  • XGBoost Regressor
  • Support Vector Regressor (SVR)
  • K-Nearest Neighbors Regressor (KNN)

Used train_test_split to divide data into 80% training and 20% testing sets.

8. Model Evaluation

Evaluated all models using:

  • Mean Squared Error (MSE)
  • Root Mean Squared Error (RMSE)
  • R² Score

9. Feature Importance

Used RandomForestRegressor.feature_importances_ to rank the importance of each input variable in prediction.

Results Summary

  • Linear Regression is overfitting (R² ≈ 1.00).
  • Random Forest and XGBoost performed well, showing robust predictions.
  • SVR and KNN showed relatively lower performance, possibly due to dataset size or scaling. Among all models Random Forest is the best model. This model is saved. ---

Solution

A Streamlit web app that helps understand and improve digital wellbeing.

Features

  • Home - Tips: Learn helpful tips for reducing screen time, improving sleep, and managing digital habits.
  • Prediction: Enter your daily habits like sleep hours, screen time, mood, and so on to get a Digital Wellbeing Score.
  • What to Do: Based on your score, get suggestions on how to improve your digital wellbeing.

Built With

Share this project:

Updates