Skip to content

LeLoUcH-V-BrItAnIa/STUDENT-S-PERFORMACE-PREDICTION-USING-MACHINE-LEARNING

Repository files navigation

🎓 Student Performance Prediction Using Machine Learning

An AI-powered educational analytics system that predicts student exam performance based on behavioral and lifestyle factors and provides personalized recommendations using Google Gemini AI.


📌 Overview

Student Performance Prediction is an EdTech project that leverages Machine Learning and Generative AI to analyze students' study habits and lifestyle patterns and predict their academic performance.

The system not only estimates exam scores but also generates personalized suggestions related to:

  • Study planning
  • Time management
  • Lifestyle balance
  • Sleep optimization
  • Mental well-being

🚀 Features

  • 📈 Student exam score prediction using Machine Learning
  • 🤖 AI-generated personalized study recommendations
  • 📊 Exploratory Data Analysis (EDA) and visualization
  • 🖥 Interactive Streamlit web application
  • 🔗 Flask backend integration
  • 🧠 Google Gemini AI assistance
  • 📉 Model comparison and evaluation

🛠 Technologies Used

Frontend

  • Streamlit

Backend

  • Flask
  • Python

Machine Learning

  • Scikit-Learn
  • Pandas
  • NumPy

Visualization

  • Matplotlib
  • Seaborn

AI Integration

  • Google Gemini API

📂 Dataset

The dataset contains approximately 1100 student records with attributes such as:

  • Age
  • Study Hours Per Day
  • Social Media Usage
  • Netflix Hours
  • Attendance Percentage
  • Sleep Hours
  • Exercise Frequency
  • Mental Health Rating

Target Variable:

Exam Score

⚙️ Machine Learning Models

Three regression models were trained and evaluated:

Model R² Score RMSE
Linear Regression 0.804 7.10
Random Forest Regressor 0.790 7.34
Decision Tree Regressor 0.720 8.47

Best Model

✅ Linear Regression

It achieved an R² score of 0.804, explaining approximately 80% of the variance in student performance.


🧠 AI Recommendation System

After prediction, the application sends the student's data to Google Gemini AI, which generates:

  • Study routine suggestions
  • Mental wellness guidance
  • Time management tips
  • Sleep optimization advice
  • Recommendations for students with part-time jobs

🏗 System Architecture

Streamlit UI
     ↓
Flask Backend
     ↓
Machine Learning Model
     ↓
Prediction Result
     ↓
Google Gemini AI
     ↓
Personalized Suggestions

📊 Project Workflow

  1. Data Collection
  2. Data Preprocessing
  3. Exploratory Data Analysis
  4. Feature Scaling
  5. Model Training
  6. Performance Evaluation
  7. Streamlit Frontend Development
  8. Flask Backend Integration
  9. Gemini AI Recommendation Generation

📸 Output

The application allows users to:

  • Input lifestyle and academic information.
  • Predict expected exam scores.
  • Receive AI-powered academic guidance.
  • Improve study habits through personalized recommendations.

Future Improvements

  • Add academic history and GPA analysis
  • Real-time progress tracking
  • Multi-semester performance prediction
  • Mental health assessment integration
  • Institution-level analytics dashboard

Author

Priyanka Kar Master of Computer Application (MCA) Swami Vivekananda University West Bengal, India


Keywords

Machine Learning • Educational Data Mining • Streamlit • Flask • Scikit-Learn • Generative AI • Google Gemini • Student Analytics • EdTech

About

This prject focuses on creating a data driven model to assess and predict the academic performace of students based factors

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors