This project predicts the selling price of a used car based on features such as year, fuel type, kilometers driven, seller type, and more.
cars.csv
— The dataset containing used car listings with features like year, present price, kms driven, fuel type, seller type, transmission, etc.Car_Price_Prediction.ipynb
— Jupyter Notebook for:- Exploratory Data Analysis (EDA)
- Data preprocessing (handling categorical features using OneHotEncoding)
- Model training using Linear Regression
- Evaluation and visualization
model.pkl
— The trained machine learning model saved usingjoblib
.ss.pkl
— StandardScaler object used to scale input features before prediction.testing_on_newdata.ipynb
— A separate notebook used to test the saved model on new data inputs.
- Model Used: Linear Regression
- Target Variable: Selling Price
- Features Used:
- Year
- Present Price
- Kms Driven
- Fuel Type (converted using one-hot encoding)
- Seller Type
- Transmission
new_data = np.array([[2018, 8.5, 25000, 0, 1, 1]]) # Encoded input
scaled = scaler.transform(new_data)
price = model.predict(scaled)
print("Predicted Price:", price)
- pandas
- numpy
- matplotlib
- seaborn
- scikit-learn
- joblib
Install them using:
pip install -r requirements.txt
-
Clone the repo:
git clone https://github.com/yourusername/Car-Price-Prediction.git
-
Open
Car_Price_Prediction.ipynb
to train the model or understand the process. -
Use
testing_on_newdata.ipynb
to test the model on unseen data.
- Venkata Sai Kumar Routhu