Spaces:
Sleeping
Sleeping
metadata
title: PriceMyCar
emoji: 🚗
colorFrom: blue
colorTo: green
sdk: docker
app_port: 7860
pinned: false
PriceMyCar
PriceMyCar is a web app to estimate used car prices in Indonesia. It uses a machine learning model to estimate the base price and adjusts it based on a 10-factor physical condition checklist.
Features
- Price Estimation: Estimates base car value based on historical sales data.
- Physical Condition Scoring: Adjusts the base price using 10 physical factors (body damage, paint, interior, accidents, flood history, etc.).
- Responsive UI: Simple UI that works on both desktop and mobile.
- Indonesian Market Adjustments: Converts the model's base currency (INR) to IDR with local inflation and brand-specific adjustments.
Local Setup
- Navigate to this directory:
cd web_app - Set up a virtual environment:
- On Windows:
python -m venv venv venv\Scripts\activate - On Mac/Linux:
python3 -m venv venv source venv/bin/activate
- On Windows:
- Install dependencies:
pip install -r requirements.txt - Run the application:
python app.py - Open
http://127.0.0.1:5000in your browser.
Pricing & Condition Scoring
The final price is calculated as:
Final Price = Base Price * (1 - Total Deduction Percentage)
The system uses 10 factors to calculate the deduction percentage:
- Body Damage Severity: 0% to -28%
- Number of Dents: -2% per dent (Max -15%)
- Paint Condition: 0% to -13%
- Interior Condition: 0% to -15%
- Accident History: 0% to -40%
- Flood Damage: 0% to -50%
- Engine & Mechanical: 0% to -30%
- Tire Condition: 0% to -5%
- Service History: +3% bonus to -6% deduction
- Modifications: 0% to -8%
Indonesian Market Adjustments
Since the base model was trained on Indian market data (in INR), we adjust it for the Indonesian market:
- Exchange Rate:
1 INR = Rp 187.6 - Inflation/Depreciation: +12% adjustment
- Market Multipliers:
- Luxury Brands (Mercedes, BMW, etc.): 1.95x
- Popular Brands (Toyota, Honda, Daihatsu, Suzuki, Mitsubishi): 1.60x
- Other Brands: 1.45x Prices are calibrated against listings on OLX Indonesia, Mobil123, and GridOto.
Model Validation
If a user inputs a brand/model that is not in the database, the system will perform a general segment approximation. A warning banner will be displayed on the result page to indicate that the prediction is an approximation.