This project is a machine learning regression system designed to analyze and predict student performance based on behavioral, academic, and lifestyle factors.
The dataset explores how study habits, digital behavior, sleep patterns, and mental health indicators influence academic outcomes.
This is a multi-output regression problem, where the model predicts multiple continuous targets such as:
- Mental Health Score
- Focus Index
- Burnout Level
- Productivity Score
- Exam Score
The dataset includes variables related to:
- Study habits
- Screen time usage
- Sleep duration
- Exercise activity
- Caffeine intake
- Demographic information
- Academic behavior
These features are used to estimate student performance indicators.
Different regression models were tested, including:
- Linear Regression
- Multi-Output Regression
- Regressor Chains
- Gradient Boosting Regressor
- Cross-validation for evaluation
Performance was measured using:
- R² Score
- Cross-validation score
- Model comparison analysis
- Data preprocessing
- Feature selection
- Train/Test split
- Model training
- Model evaluation
- Multi-target prediction
- Python
- Pandas
- NumPy
- Scikit-learn
- Machine Learning Techniques
The objective of this project is to understand how lifestyle and academic factors influence student performance and to build a predictive system capable of estimating multiple performance indicators simultaneously.
- Advanced feature engineering
- Hyperparameter tuning
- Non-linear models (e.g., boosting techniques)
- Neural network implementation
- Deployment as a web application
This project was developed for learning and experimentation in multi-output regression and real-world data analysis.