Carvana - Predict Car Prices

Overview

This dataset contains data about 22,000 cars from carvana.com. Using this data we can derive new insights about cars prices, and we can develop predictive models to predict the cost of car given it's make and model, year, and mileage.

Tasks

• Create a predictive model to calculate the price of the car.
• Using natural language processing, create an algorithm to find a car's price given it name (make and model).
• Is there a correlation/relationship between between the year of a car and a car's price?
• Is there a correlation/relationship between between the mileage on a car and a car's price?

Data Details

• The data was web scraped from carvana.com/cars.
• There are 22,000 rows of data (first 1000 pages of the site).
• The name column contains the make and model of the car.
• The cars are used and the miles column says the number of miles the car has already driven.
• The price of the car is in USD.