Car Purchase Prediction Model🚘

This project implements a Gradient Boosting Classifier to predict car purchase likelihood using a Japanese dataset and applies the model to estimate potential customers in an Indian dataset. The Jupyter Notebook (Car_Purchase_Prediction.ipynb) includes data preprocessing, feature engineering, model training, evaluation, business insights, and a Tableau visualization plan.

Dataset

• Japanese Dataset: JPN Data.xlsx - CN_Mobiles.csv
• Indian Dataset: IN_Data.xlsx - IN_Mobiles.csv

Installation

pip install pandas sklearn seaborn matplotlib

• Requires Python 3.8+ and Jupyter Notebook.
• Place dataset files in the project directory.

Usage

1. Clone the repository:

   git clone https://github.com/itsbk13/Car_Purchase_Prediction.git
   cd Car_Purchase_Prediction

2. Run Car_Purchase_Prediction.ipynb in Jupyter to execute the full workflow.
3. Use exported japan_data.csv and india_data.csv for Tableau visualizations.

Project Structure

Car_Purchase_Prediction/
├── JPN Data.xlsx - CN_Mobiles.csv    # Japanese dataset
├── IN_Data.xlsx - IN_Mobiles.csv     # Indian dataset
├── Car_Purchase_Prediction.ipynb     # Main notebook
├── japan_data.csv                   # Exported Japanese data
├── india_data.csv                   # Exported Indian data
├── output.csv                       # Indian predictions
└── README.md                        # This file

Key Features

• Model: Gradient Boosting Classifier (70.26% accuracy).
• Key Predictors: Car age (>360 days) and income.
• Business Insight: Target high-income individuals with older cars.
• Tableau: Visualize age, income, and purchase trends.

License

• This project is licensed under the MIT License.
• Attribution: This project was built as a capstone during an internship program with Internshala.