This project contains the detailed overview of used car price prediction using data mining algorithms and methods. Various regression models have been explored including simple multiple linear regression, LASSO linear regression, random forest regression and multilayer perceptron, among which a random forest model provides the highest predicting power with R2 = 0.8764 on the test dataset. The notebook file with all the code to run the models and generate the figures and outputs is uploaded public on Kaggle.
Data source: CARS DATASET (Audi, BMW, Ford, Hyundai, Skoda, VW).
The notebook file with all the code to run the models and generate the figures and outputs is also uploaded public on Kaggle: Used Car Price Prediction.