CAR PRICE PREDICTION USING LINEAR REGRESSION

by Igweze Ifeanyi

Dataset

This dataset contains information about used cars listed on different websites of shape 301 x 8 and 2 duplicated rows.

Objective

Create a good Linear Regression model to predict the selling prices of the test data.

Summary of Findings & Actions

The two main features with the highest correlation with the selling price where the Present Price and the Fuel Type Petrol (Gotten after encoding the Fuel Type column). Before creating the model I dropped the Car Name and Owner columns.

Key Insights

From this, I created a Linear Reggression model with R2 score of 87.04% in the main branch, But in the Dropped-duplicates branch, I dropped the duplicated rows and had my R2 score drop to 75.81%