Car price prediction

Description :-

• This project aims to estimate a car's price based on its features using machine learning techniques. The dataset used in this project can be found here.

Data Cleaning :-

• Eliminating the duplicate columns, replacing a specific column that contains a special character with nan values
• The outliers were removed using the iqr approach.

Feature Engineering :-

• Filling in a missing value with the KNN Imputer
• Instead of utilising ordinal encoding, because some of the columns contained more than 20 categorical variables. Target guided encoding that I utilised provides greater correlation with the dependent variable.

Model Creation :-

Decision Tree Regressor

• Beginning with a decision tree regressor, training the input data with a decision tree method and model worked extremely well and produced training data with a high degree of accuracy. The test data's predictions were correct.

Model Evaluation :-

• The R2 score achieved was 0.99, indicating a highly accurate model. The mean absolute error was 17.79, and the mean absolute percentage error was 0.005.

Conclusion :-

Model provided good accuracy, thus I didn't even adjust the hyperparameters. Target directed encoding and mean encoding were used in place of label encoding, and they produced substantial correlations with the target variable.