Marg Darshan

Overview

Marg Darshan is a Flask-based career recommendation web application that predicts a suitable job profile for a student based on their academic scores, problem-solving and creativity ratings, hackathon experience, and top technical skills.

The app combines several machine learning models saved in the models/ directory and uses majority voting to provide a final suggestion.

What it does

• Hosts a web interface to collect student performance data
• Sends the input to a Flask /predict endpoint
• Loads trained models from models/
• Returns one or more predicted job profiles

Supported prediction models

The project loads and uses the following models:

• Decision Tree
• Random Forest
• Linear Regression
• K-Nearest Neighbors (KNN)
• Support Vector Machine (SVM)

Note: The gradient boosting model file exists in models/gradient_boosting_model.jolib, but its prediction is currently commented out in pred.py.

Expected input

The interface collects:

• Academic marks for:
  • Data Structure
  • Database management
  • Operating System
  • Computer Network
  • Mathematics
  • Aptitude
  • Communication Skills
• Hackathon count
• Problem solving rating (1-10)
• Creativity rating (1-10)
• Top 2 selected skills from the available options

Project structure

• app.py - Flask web server and /predict route
• pred.py - prediction logic, model loading, input encoding, and final voting
• requirements.txt - Python dependencies
• templates/layout.html - front-end HTML template
• static/JS/app.js - client logic for collecting form data and calling the API
• static/CSS/main.css - front-end styling
• models/ - saved machine learning models used for prediction

Installation

1. Create a virtual environment (recommended):

   python3 -m venv venv
   source venv/bin/activate

2. Install dependencies:

   pip install -r requirements.txt

Running the app

Start the Flask server:

python app.py

Then open the browser at:

http://127.0.0.1:5000

How to use

1. Enter your subject marks and hackathon count.
2. Rate your problem solving and creativity skills.
3. Select exactly two skills from the skill checklist.
4. Submit the form to receive the recommended job profile(s).

Notes

• The app currently uses model files loaded from models/.
• Predictions are returned as a JSON array of career labels.
• If the UI or API behavior needs improvement, the main frontend flow is in static/JS/app.js and the backend logic is in pred.py.

Dependencies

Built with:

• Flask
• pandas
• numpy
• scikit-learn
• joblib
• scipy
• threadpoolctl
• gunicorn