idrissalk/Car-Price-Prediction-

A machine learning project that predicts the fair market price of used cars in Morocco based on features such as brand, model, year, mileage, and more. Built with Python, Flask, and scikit-learn, this project provides a web interface for interactive price predictions. # 🚗 Car Price Prediction - Morocco This project aims to build a **Machine Learning application** to predict the price of used cars in Morocco. The goal is to provide a **reliable and accurate estimate** based on various features such as brand, model, year, mileage, fuel type, and more. --- ## Introduction The used car market in Morocco is growing rapidly, but prices often lack structure and depend on negotiation, location, and seller perception. This project aims to: - Identify the **key factors** that influence car prices. - Provide an **automatic and reliable price estimation**. - Help buyers, sellers, and platforms **make data-driven decisions**. **Expected result:** An easy-to-use web application that predicts the price of a used car based on its specifications. --- ## Dataset The dataset contains **24,776 real ads** collected from the Moroccan market via Avito. After cleaning and processing, the final dataset contains **20,324 records** and **28 features**. **Main features include:** - **Brand** (e.g., Volkswagen, Renault, Dacia, Toyota, etc.) - **Model** - **Year of manufacture** - **Mileage** - **Fuel type** (Diesel, Gasoline, Hybrid) - **Fiscal power** - **Transmission type** (Manual, Automatic) - **Vehicle condition** - **Price** --- ## Data Preprocessing ### Main steps: 1. **Handling missing values** and replacing anomalies. 2. Transforming mileage from ranges (`"200000 - 249999" → 224999.5`). 3. **Encoding categorical variables** using `LabelEncoder` and `OneHotEncoder`. 4. Scaling numerical variables using `StandardScaler`. 5. Removing duplicates and irrelevant data (e.g., extreme prices, rare brands). --- ## Exploratory Data Analysis (EDA) Some visualizations were created to better understand the data: - Distribution of cars by **brand**. - Average price by **brand** and **fuel type**. - Impact of **mileage on price**. - **Correlation matrix** between numerical variables. --- ## Modeling Several machine learning algorithms were …