Kodluyoruz Izmir Data Science Bootcamp Capstone Project
We use dataset from Kaggle for used car price prediction.
Dataset : https://www.kaggle.com/austinreese/craigslist-carstrucks-data
We tried to predict the price range of used cars given the features by doing multi-class classification.
First we discarded features that are not relevant for the prediction of the price ranges. And then dropped some NA values and removed outliers.
Secondly, we encoded data with label encoding and ordinal encoding methods.
And then, nan values of cylinders, size, type, drive features filled with Random Forest model.
Third, we made six price ranges, which means six price range classes.
Finally, we applied random forest, xgboost, gradient boosting and decision tree models on the preprocessed dataset with features as inputs and the price ranges as output.
We also used GridSearchCV to find best parameters for our models.