VisionED is an AI-powered Student Performance Prediction System that uses Machine Learning (Random Forest Regressor) integrated into a Flask-based web application.
The system predicts End-Semester Marks and Attendance based on students' academic performance indicators such as previous marks and attendance.
This project was developed as a Minor Project for New Government Polytechnic, Patna-13 (NGP).
VisionED provides a structured academic platform where students and administrators can manage academic analytics, study materials, announcements, and academic queries.
- AI-based prediction of End-Semester Marks and Attendance
- Branch & Semester specific academic analysis
- Profile completion-based secure dashboard access
- Structured CSV-based academic data upload system
- Study material upload and management by Admin
- Announcement management system
- Academic query submission and Admin response system
- Role-based access control (Student / Admin / Super Admin)
- Student performance analytics dashboard
Landing page with role-based login and signup system.
Students can access prediction analysis, study materials, announcements, and academic queries.
Administrative control panel used to manage users, upload academic data, and train prediction models.
Admin selects Branch and Semester, downloads the structured template, fills previous batch data (marks & attendance), and uploads it to train the model.
Subject-wise prediction results generated using the Random Forest Regressor model.
Detailed analytics insights based on student academic inputs.
Admin interface used to manage students and other administrators.
- Python
- Flask
- Flask-SQLAlchemy
- Werkzeug
- Scikit-learn (Random Forest Regressor)
- Pandas
- NumPy
- OpenPyXL
- HTML5
- CSS3
- Bootstrap 5
- JavaScript
- SQLite
The system uses Scikit-learn’s Random Forest Regressor to model nonlinear relationships between academic indicators and final end-semester outcomes.
- Previous Semester Marks
- Previous Semester Attendance Percentage
- Internal Subject Marks (entered by students during prediction)
- Final End-Semester Marks
- Final End-Semester Attendance
The model is trained separately for each Branch and Semester using structured historical academic data from previous student batches.
Once trained, the system predicts expected end-semester marks and attendance based on student academic inputs.
VisionED-performance-predictor/
│
├── static/
│ └── uploads/
│ └── images/
│ ├── admin_default.png
│ ├── student_default.png
│ └── profile images...
│
├── templates/
│ ├── admin_dashboard.html
│ ├── admin_material_uploader.html
│ ├── admin_announcements.html
│ ├── registered_users.html
│ ├── student_dashboard.html
│ ├── student_analytics.html
│ ├── profile_student.html
│ ├── profile_admin.html
│ ├── login.html
│ ├── signup.html
│ ├── index.html
│ ├── contact.html
│ ├── blog.html
│ ├── faq.html
│ ├── privacy.html
│ ├── team.html
│ └── other templates...
│
├── app.py
├── requirements.txt
└── README.md
git clone https://github.com/sagarcs818/VisionED-performance-predictor.git
cd VisionED-performance-predictorpython -m venv venv
venv\Scripts\activatepython3 -m venv venv
source venv/bin/activatepip install -r requirements.txtOpen app.py and update:
app.secret_key = "your_strong_secret_key_here"🔐 This key is used to:
- Secure login sessions
- Protect admin authentication
- Prevent session tampering
- Enable flash messages
Replace it with a strong random string.
⚠ If this key changes later, all users will be logged out automatically.
- Admin Signup Code:
1234 - Super Admin Code:
5678
These can be changed later from the Admin Profile Page.
python app.pyhttp://127.0.0.1:5000/
The following diagram illustrates how users navigate through the system and how workflows are connected:
- Register using the provided Admin Code
- Login to the system
- Complete required profile details
- Access the Admin Dashboard
⚠ Profile completion is required to access dashboard features.
- From the Admin Dashboard, navigate to Material Uploader
- Select Upload Analytics Data
- Choose:
- Branch
- Semester
- Download the structured CSV template generated by the system
The template includes required columns such as:
- Previous Semester Marks
- Attendance Percentage
- Final End-Semester Marks
- Fill the CSV template with historical student data
- Ensure the column structure is correct
- Upload the completed CSV file
- The system trains the Random Forest model
Once uploaded successfully, the system becomes ready to generate predictions for students.
After profile completion, Administrators can:
- 📂 Upload study materials
- 📢 Post announcements
- ❓ Respond to student queries
- 👥 Manage registered users
- 📊 View student prediction analytics
- 🛡 Super Admin privileges for managing other admins
- Register as a Student
- Login to the system
- Complete academic profile:
- Branch
- Semester
⚠ Profile completion is required to access dashboard features.
Navigate to Performance Analysis and enter:
- Internal Subject Marks
- Previous Semester Marks
- Previous Semester Attendance Percentage
- Submit academic inputs
- The system runs the trained Random Forest model
- Predicted End-Semester Marks and Attendance are displayed
After profile completion, Students can:
- 📊 Perform performance analysis
- 📂 View & download study materials
- 📢 View announcements
- ❓ Submit academic queries
- 👍 Interact with posts (Like / Reply)
VisionED aims to:
- Identify students at academic risk early
- Provide AI-based academic insights
- Improve academic decision making
- Digitize academic analytics in institutions
⭐ If you found this project useful, consider starring the repository.
This project is licensed under the MIT License.
See the LICENSE file for more information.