STUDENT PERFORMANCE PREDICTOR

Project Overview

This project analyzes how various factors influence student performance in mathematics. Using machine learning algorithms, it predicts math scores based on:

• Demographics: Gender, race/ethnicity
• Socioeconomic factors: Parental education level, lunch type
• Academic preparation: Test preparation course completion
• Prior performance: Reading and writing scores

The application provides a user-friendly web interface for making predictions.

Dataset Information

• Source: Kaggle - Students Performance in Exams
• Size: 1,000 records with 8 features

Key Features:

Feature	Description
gender	Student's gender (male/female)
race_ethnicity	Ethnic group (Group A-E)
parental_level_of_education	Parent's highest education level
lunch	Lunch type (standard/free or reduced)
test_preparation_course	Test prep completion (completed/none)
reading_score	Reading test score (0-100)
writing_score	Writing test score (0-100)
math_score	Target variable - Mathematics test score (0-100)

Model Workflow

This project follows a structured machine learning workflow to predict student math scores from start to deployment.

1. Data Ingestion & Preprocessing

   • Load dataset and split into training (80%) and test (20%) sets.
   • Handle missing values and perform feature engineering:
     • Numerical features are scaled using StandardScaler.
     • Categorical features are encoded with OneHotEncoder and scaled.

2. Model Training & Selection

   • Train multiple regression models: Linear Regression, Lasso, Ridge, K-Neighbors, Decision Tree, Random Forest, XGBoost, CatBoost, AdaBoost, Gradient Boosting.
   • Perform hyperparameter tuning using GridSearchCV.
   • Evaluate models using R², MAE, and RMSE on both training and test sets.
   • Select the best-performing model (Ridge Regression) based on test set performance.

3. Model Deployment

   • Serialize the trained model and preprocessing pipeline using pickle and dill.
   • Serve predictions through a Flask web application with a responsive UI built using HTML, Tailwind CSS, and JavaScript.
   • Deploy the application on Render for public access.

Performance Metrics

The project evaluates multiple regression models to predict student math scores. The table below summarizes R², Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE) for both training and test sets.

Model	Train R²	Test R²	Train MAE	Test MAE	Train RMSE	Test RMSE
Linear Regression	0.874	0.880	4.278	4.228	5.327	5.411
Lasso Regression	0.807	0.825	5.206	5.158	6.594	6.520
Ridge Regression	0.881	0.881	4.156	4.102	5.231	5.198
K-Neighbors Regressor	0.784	0.784	4.980	5.022	6.215	6.310
Decision Tree Regressor	0.722	0.722	5.670	5.712	7.102	7.150
Random Forest Regressor	0.855	0.855	3.912	4.020	4.892	4.950
XGBoost Regressor	0.848	0.828	4.102	4.210	5.021	5.112
CatBoost Regressor	0.860	0.852	3.980	4.052	4.932	5.021
AdaBoost Regressor	0.842	0.845	4.102	4.150	5.120	5.198
Gradient Boosting	0.845	0.846	4.050	4.102	5.050	5.098

Best Model: Ridge Regression

• Achieves the highest R² on the test set (0.881)
• Provides reliable predictions with low MAE (~4.1) and RMSE (~5.2 points)

Instructions to Use

Web Interface

• Home Page: Overview and project information
• Prediction Form: Enter student details:
  • Select gender and ethnicity
  • Choose parental education level
  • Specify lunch type and test preparation
  • Input reading and writing scores (0-100)
• Results: Get instant math score prediction

Project Structure

student-performance-predictor/
├── app.py                  # Flask web application
├── requirements.txt        # Python dependencies
├── setup.py                # Package setup
├── Procfile                # Deployment config
├── README.md               # Project documentation
│
├── src/                    # Source code
│   ├── __init__.py
│   ├── exception.py        # Custom exception handling
│   ├── logger.py           # Logging configuration
│   ├── utils.py            # Utility functions
│   │
│   ├── components/         # ML pipeline components
│   │   ├── data_ingestion.py
│   │   ├── data_transformation.py
│   │   └── model_trainer.py
│   │
│   └── pipeline/           # Inference pipelines
│       ├── train_pipeline.py
│       └── predict_pipeline.py
│
├── templates/              # HTML templates
│   ├── base.html
│   ├── index.html
│   └── home.html
│
├── artifacts/              # Model artifacts
│   ├── model.pkl
│   ├── preprocessor.pkl
│   ├── train.csv
│   ├── test.csv
│   └── data.csv
│
└── notebook/
    ├── 1_EDA_Student_Performance.ipynb
    ├── 2_Model_Training.ipynb
    └── data/
        └── stud.csv

Key Technologies & Libraries

• Backend: Flask, Python 3.8+
• Machine Learning & Modeling: Scikit-learn, CatBoost, XGBoost, Lasso, Ridge, AdaBoost
• Data Processing: Pandas, NumPy
• Visualization: Matplotlib, Seaborn, Plotly
• Frontend: HTML5, Tailwind CSS
• Deployment: Render
• Serialization: Pickle, Dill

Installation

Prerequisites

• Python 3.8 or higher
• pip package manager

Local Setup

• Clone the repository

git clone https://github.com/yourusername/student-performance-predictor.git
cd student-performance-predictor

• Create virtual environment

python -m venv venv
source venv/bin/activate   # Windows: venv\Scripts\activate

• Install dependencies

pip install -r requirements.txt

• Run the application

python app.py

• Open browser at http://localhost:5000

---