Summary

Introduction

I have made prediction model for used car price for the competition held in Dacon(Usually known as Korean version of Kaggle)

Competition Link: https://dacon.io/en/competitions/official/235901/data

I used CatBoost algorithm for the prediction model and implemented various preprocessing methods to maximize the model’s accuracy.

Datasets

Train Data → 10 X variables related to each car and sales price (y variable). Total 1,015 rows.

Test Data → 10 X variables related to each unit of cars. Total 436 rows.

Language and libraries

Python, Pandas, Sklearn, Seaborn

Data Preprocessing

Outliers

I found 2 outliers in ‘year’ columns having values around 1200 and deleted them.

Outliers in Year column. They are deleted in the dataset

Imputation

No missing value was found in both train and test dataset