Project link: https://car-price-price.herokuapp.com Demo Video: https://youtu.be/HEaFU68WAPM
This project aims to predict the Price of an used Car by taking it's Company name, it's Model name, Year of Purchase, and other parameters.
- Clone the repository
- Install the required packages in "requirements.txt" file.
Some packages are:
- numpy
- pandas
- scikit-learn
- Run the "application.py" file And you are good to go.
- This project takes the parameters of an used car like: Company name, Model name, Year of Purchase, Fuel Type and Number of Kilometers it has been driven.
- It then predicts the possible price of the car. For example, the image below shows the predicted price of our Hyundai Grand i10.
-
First of all the data was scraped from Quikr.com (https://quikr.com) Link for data: https://github.com/rajtilakls2510/car_price_predictor/blob/master/quikr_car.csv
-
The data was cleaned (it was super unclean :( ) and analysed.
-
Then a Linear Regression model was built on top of it which had 0.92 R2_score.
Link for notebook: https://github.com/rajtilakls2510/car_price_predictor/blob/master/Quikr%20Analysis.ipynb
- This project was given the form of an website built on Flask where we used the Linear Regression model to perform predictions.