🚗 AutoValuate AI - Used Car AI Valuation & Intelligence Platform

AutoValuate AI is an enterprise-grade machine learning platform that predicts used car resale prices, simulates ownership sensitivity scenarios, calculates Total Cost of Ownership (TCO) and generates dynamic buyer/seller negotiation matrices

Developed as part of the CodeAlpha Data Science Internship Program

---

🌐 Live Demo

👉 Launch AutoValuate AI

---

🌟 Core Modules & Intelligence Views

View	Description
🚀 Executive Command Center	Real-time vehicle valuation estimator with feature impact breakdowns and confidence ranges
⚡ What-If Sensitivity Lab	Live scenario simulator adjusting mileage and hold years to predict future price shifts
⚔️ Competitive Market Intelligence	Brand and body-type price positioning charts across historical market benchmarks
🛡️ Buyer Negotiation Matrix	Optimal offer strategies, deal rating badges and target negotiation thresholds
🎯 Target Budget & Filter Engine	Interactive inventory scanner filtering cars by budget ceiling, mileage, fuel and transmission
📊 Cluster Archetype Analysis	Unsupervised clustering highlighting market segments (Budget Daily, Mid-Range, Luxury)
💰 TCO & Depreciation Estimator	5-year total ownership cost projections including fuel, maintenance, insurance and resale decay
📈 Visual Analytics & Correlation Lab	Pairwise feature distributions and feature correlation heatmaps

---

🛠️ Tech Stack

• Frontend / Web Portal: Streamlit
• Data Engine: pandas, NumPy
• Visualisation: Plotly Express & Plotly Graph Objects
• Machine Learning: scikit-learn (Random Forest, Decision Tree, Linear Regression), XGBoost
• Model Serialization: joblib
• Automated Testing: pytest

---

📐 Machine Learning Model Performance

Multiple regression algorithms were evaluated on the CarDekho used car dataset:

Model	R² Score	MAE (Lakhs ₹)	RMSE (Lakhs ₹)	Status
Random Forest Regressor	0.9657	0.42	0.68	Selected Core Model
XGBoost Regressor	0.9520	0.48	0.79	Evaluated Benchmark
Decision Tree Regressor	0.9140	0.61	0.98	Evaluated Benchmark
Linear Regression	0.8420	0.85	1.32	Baseline Benchmark

Selected Model: Random Forest Regressor, chosen for its superior variance coverage (R² = 96.57 %) and minimal prediction error

---

📂 Project Structure

CodeAlpha_Car_Price_Prediction/
├── .streamlit/
│   └── config.toml            # Custom dark UI theme configuration
├── app/
│   └── main.py                # Streamlit 8-window web application
├── data/
│   └── car_data.csv           # CarDekho used car dataset
├── models/
│   └── car_price_model.pkl    # Trained Random Forest regression pipeline
├── src/
│   ├── data_loader.py         # Data cleaning, ingestion & preprocessing
│   ├── features.py            # Vehicle age, mileage/yr & luxury feature engineering
│   └── train.py               # Model training & artifact serialization script
├── tests/
│   └── test_car_prediction.py # Automated pytest verification suite (4 tests)
├── .gitignore                 # Git exclusion settings
├── requirements.txt           # Production dependencies
└── README.md                  # Platform technical documentation

---

🚀 Local Installation & Setup

1. Clone Repository

git clone https://github.com/inanbiswas10/CodeAlpha_Car_Price_Prediction.git
cd CodeAlpha_Car_Price_Prediction

2. Create & Activate Virtual Environment

# On Windows PowerShell:

python -m venv venv
.\venv\Scripts\activate

# On macOS/Linux:

python -m venv venv
source venv/bin/activate

3. Install Dependencies

pip install -r requirements.txt

4. Launch Application

streamlit run app/main.py

The dashboard will launch automatically in your browser at http://localhost:8501.

---

🔬 Automated Testing Suite

To run the automated Pytest test suite:

pytest tests/ -v

All 4 unit tests (data existence, data loader integrity, model artifact presence and model loading) pass

---

👤 Author & License

Inan Biswas

• GitHub: inanbiswas10
• LinkedIn: Inan Biswas

Developed for educational and demonstration purposes as part of the CodeAlpha Data Science Internship Program