This project aims to predict the prices of used cars based on various features such as model, year, and mileage. The project involves comprehensive analysis of car data, selection of suitable machine learning models, and accurate prediction of car prices using these models.
Train Data Set: This data set includes information about used cars such as model, year, mileage, engine capacity, power, and price.
A data set related to used car prices was selected and used.
Data cleaning, handling missing values, normalization, and other preprocessing steps were performed.
- Linear Regression: Used as a basic regression model.
- Ridge Regression: Used as a regularized version of linear regression.
- Lasso Regression: Used for feature selection.
- ElasticNet Regression: Used as a combination of Ridge and Lasso regressions.
- K-Nearest Neighbors (KNN) Regression: Used as a simple machine learning algorithm.
Each model was trained on the training data and evaluated using performance metrics such as accuracy, F1 score, and ROC AUC score.
The performance metrics obtained from the project demonstrated the effectiveness of various regression models in predicting car prices. In particular, certain models provided higher accuracy and better F1 scores compared to others.
For more information or questions related to the project, you can contact me:
Emirhan Tozlu
LinkedIn Profile