Auto Price Prediction: Data Analysis and Predictive Modeling

---

Table of Contents

• Project Overview
• Project Structure
• Key Features
• Installation
• Usage
• Results
• Challenges & Solutions
• Contributing
• License
• Contact

---

Project Overview

Objective: Build a predictive model to understand how car prices vary with design and engineering features, enabling strategic adjustments in automotive design and pricing.

Dataset: 1985 Auto Imports Database with 205 instances and 26 attributes.

Key Tasks:

1. Data Analysis: Clean, explore, and preprocess data.
2. Predictive Modeling: Train and compare regression models.
3. Insight Generation: Translate model results into actionable business strategies.

---

Project Structure

├── data
│   ├── 1.1 raw               # Raw datasets (auto_imports.csv)
│   └── 1.2 processed         # Processed training/testing splits
├── docs                      # Problem statement and documentation
├── notebooks                 # Jupyter notebooks for analysis and modeling
├── report                    # Final Report.md and supporting files
├── results
│   ├── 365csv pre-analysis   # EDA outputs (statistics, visualizations)
│   ├── figures               # Saved plots (boxplots, histograms)
│   └── models                # Serialized models (final_lasso_model.joblib)
├── scripts                   # Utility scripts (AutoPrice.py, utility.py)
├── LICENSE
├── README.md
└── requirements.txt          # Python dependencies

---

Key Features

• Data Preprocessing:

  • Missing value imputation (median/mode).
  • Outlier capping using IQR and domain knowledge.
  • PCA for dimensionality reduction (95% variance retained).

• Model Development:

  • Compared 7 models: Lasso, XGBoost, Gradient Boosting, etc.
  • Hyperparameter tuning with cross-validation.

• Deployment-Ready:

  • Best model: Lasso Regression (R² = 0.917, RMSE = 1,987).
  • Interpretable coefficients for business strategy (e.g., BMW adds $7,347 to price).

---

Installation

1. Clone the repository:

   git clone https://github.com/dhaneshbb/AutoPricePred.git
   cd AutoPricePred

2. Install dependencies:

   pip install -r requirements.txt

3. Launch Jupyter Notebook:

   jupyter notebook

---

Usage

1. Data Analysis:

   • Run AutoPricePred_Analysis.ipynb to explore data cleaning, EDA, and visualizations.

2. Model Training:

   • The notebook includes code for model comparison, hyperparameter tuning, and saving the best model.

3. Inference:

   import joblib
   model = joblib.load('models/final_lasso_model.joblib')
   prediction = model.predict([input_features])

---

Results

Model	Test RMSE	Test R²	Training Time (s)	Overfit (Δ R²)
Lasso (α=10)	1,987	0.917	0.019	0.033
XGBoost	1,663	0.942	0.143	0.052
Gradient Boosting	1,842	0.928	0.118	0.051

Final Model: Lasso Regression (α=10)

• Rationale: Balances interpretability, speed, and generalizability.
• Key Drivers: Luxury brands (make_bmw, make_mercedes-benz), engine location, and PCA components.
• Performance:

Metric	Value
Test R²	0.917
Test RMSE	1,987
Cross-Validation R²	0.899 ± 0.027

Key Insights:

• Luxury brands (BMW, Mercedes) command significant price premiums.
• Rear-engine vehicles are associated with higher prices.
• Vehicle size/power (PCA_1) is a critical pricing factor.

---

Challenges & Solutions

Challenge	Solution
Missing Values (18%) with data Leakage	Median/mode imputation + column drop.
Multicollinearity	PCA and VIF-based feature removal.
High Cardinality	Regularization (Lasso) for sparsity.
Model Overfitting	Cross-validation and hyperparameter tuning.

---

Contributing

Contributions are welcome!

1. Fork the repository.
2. Create a feature branch: git checkout -b feature/new-feature.
3. Commit changes: git commit -m 'Add new feature'.
4. Push to the branch: git push origin feature/new-feature.
5. Submit a pull request.

---

License

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

---

Made with using insightfulpy