Regression of Used Car Prices

This repository contains a machine learning project aimed at predicting the prices of used cars based on various features. The project leverages data preprocessing, feature engineering, and regression modeling to deliver accurate price predictions.

Project Overview

The primary goal of this project is to develop a predictive model for used car prices using a variety of regression techniques. The dataset includes various features such as car brand, model, year, mileage, fuel type, and more.

Models Used

In this project, multiple regression models were implemented and evaluated, including:

1. Gradient Boosting Regressor (GBR): The best-performing model, which outperformed all other regression techniques in predicting used car prices.
2. Neural Network Regressor (MLPRegressor): A feedforward neural network used for regression, with hyperparameter tuning for optimization.
3. Linear Regression: Implemented using cuML for GPU acceleration to evaluate its performance on the dataset.
4. Lasso Regression: Utilized with hyperparameter tuning to improve performance and assess feature importance.
5. Random Forest Regressor: Employed for its ensemble learning capabilities and feature importance analysis.
6. Support Vector Machine (SVM): Applied for regression tasks to explore its predictive power.
7. K-Nearest Neighbors (KNN): Used as a comparison model to evaluate performance against other algorithms.

Key Findings

• The Gradient Boosting Regressor (GBR) consistently outperformed other models, demonstrating superior accuracy in predicting used car prices.
• Hyperparameter tuning was crucial for optimizing model performance, particularly for the neural network and Lasso regression models.

Getting Started

1. Clone the repository:

   git clone https://github.com/yourusername/Regression-of-Used-Car-Prices.git
   
   cd Regression-of-Used-Car-Prices

2. Install the required packages:

pip install -r requirements.txt

Usage

Prepare your dataset and place it in the specified directory. Run the Jupyter Notebook (Used_Car_Price_Prediction.ipynb) to execute the data preprocessing, model training, and prediction steps. The final predictions will be saved in a CSV file for submission.

Contributing

Contributions are welcome! If you have suggestions for improvements or new features, feel free to fork the repository and submit a pull request.

License

This project is licensed under the MIT License. See the LICENSE file for details.

You can copy and paste this into your README.md file in your GitHub repository. Make sure to replace yourusername with your actual GitHub username in the installation section.