The problem carzoo solves

This problem statement belongs to the data analytics domain and aims to leverage machine learning techniques to develop a predictive model that can accurately predict the price of used cars. The large dataset provided includes information about thousands of used cars, including make, model, year, mileage, condition, and price. The goal is to develop a predictive model that can accurately predict the price of a used car based on these and other factors.

Challenges we ran into

In the dataset for the locations column, we were initially trying to one-hot encode and drop the first column to transform the categorically data. This would work for cases when the user enters location value on which the model has been trained. The similar issue was also faced for the Fuel Type column. We overcame this by not dropping the first column after one-hot encoding. If a user enters a new location value then all column values will be 0 and the model won't throw an error. This was one of the few issues that we faced earlier but overcame.