AI-Enabled Used Cars Price Predictor

Overview

This project focuses on predicting the prices of used cars in Pakistan using an AI-based approach. The data for training the model was scraped from PakWheels, a popular website dedicated to selling cars in Pakistan. The web scraping was performed using a Python spider.

Features

• Web Scraping: Utilized Python spider to collect data from the PakWheels website.
• Model Training: Employed GradientBoostingRegressor for training the predictive model.
• Model Accuracy: Achieved an impressive accuracy of approximately 90% during training.
• Model Persistence: The trained model has been saved for later use.
• GUI with Flask: Implemented a user-friendly web interface using Flask for users to input information and receive price predictions.

Requirements

• Python 3.x
• Flask
• scikit-learn

Usage

1. Clone the repository:

   git clone https://github.com/hzaheer48/used-cars-price-predictor.git
   cd used-cars-price-predictor
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Run the Flask app:

   python app.py
   

4. Open your browser and navigate to http://localhost:5000 to use the application.

Model Training

The model training process is detailed in the model_training.ipynb notebook. For optimal results, consider retraining the model with updated data.

Screenshots

Contributions

Feel free to contribute to this project by opening issues or creating pull requests.