Student Performance Prediction – README

This project analyzes student performance data and builds various machine learning regression models to predict the Performance Index based on multiple academic and personal factors.

---

📌 Project Overview

This notebook performs complete end‑to‑end data analysis and regression modeling using Python. It includes:

• Data loading and inspection
• Exploratory data analysis (EDA)
• Visualization
• Outlier detection
• Feature scaling and encoding
• Train/validation/test splits
• PCA transformation
• Multiple regression algorithms
• Evaluation using MSE
• Visualization of predictions

---

📁 Dataset

The dataset used:

Student_Performance.csv

It includes various features such as:

• Hours Studied
• Previous Scores
• Attendance
• Sleep Hours
• Extracurricular Activities
• ...and more

The target variable is:

• Performance Index

---

🚀 Workflow Summary

1. Import Necessary Libraries

The project uses libraries like NumPy, Pandas, Matplotlib, Seaborn, Scikit-Learn, and XGBoost.

2. Load and Inspect Data

• Display shape, head, and info
• Check for missing values

3. Visualizations & EDA

• Box plots for outlier detection
• Histograms for distribution
• KDE plots
• Pie chart for categorical distribution
• Pairplots for feature interaction
• Correlation heatmap

4. Data Cleaning

• Remove duplicates
• Handle numerical scaling with MinMaxScaler
• Label encode categorical features

5. Train / Validation / Test Split

• 70% training
• 30% temporary → split into validation and test (70/30)

6. PCA Transformation

Principal Component Analysis is applied to validation and test sets.

7. Regression Models Used

The notebook trains and evaluates:

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

Each model outputs:

• Validation MSE
• Test MSE
• Prediction vs Actual comparison table
• Scatter plot of predicted vs actual values

8. Model Evaluation Metrics

• Mean Squared Error (MSE)
• Visualization of predictions

A helper function (plot_predicted_vs_actual_seaborn) is included to show model performance.

---

📊 Visualization

Several plots are included to understand:

• Feature distribution
• Outliers
• Relationships between variables
• Model prediction accuracy
• Correlation heatmap

---

🧠 Goal of the Project

To identify:

• Which factors impact student performance
• Which regression model provides the best prediction accuracy

---

⚙️ Technologies Used

• Python 3
• NumPy, Pandas – Data handling
• Matplotlib, Seaborn – Visualization
• Scikit-Learn – ML models & preprocessing
• XGBoost – Gradient boosting model

---

📈 Expected Outcomes

• A clear comparison of multiple regression models
• Visual insights into student behavior and its influence on performance
• Clean and preprocessed dataset ready for model training

---

📦 How to Run the Notebook

1. Install dependencies:

   pip install numpy pandas matplotlib seaborn scikit-learn xgboost

2. Download the dataset into the working directory.

3. Run all notebook cells sequentially.

---

📝 Author

This project was created as part of a machine learning study and demonstrates full model development workflow.

---

If you need help improving the notebook, optimizing models, or creating visual summaries, feel free to ask!