In the second and third week of the Metis Data Science Bootcamp, every participant choose and embark on his personal regression machine learning project. In my project, I webscraped (with BeautifulSoup) around 20,000 footballers' data from Fifa Index and attempt to use their personal metrics and skills ratings to predict their market value. Prediction of market value is important as soccer clubs spend astronomical amount of money on footballer acquisition.
Through this project, I employ skills such as feature selection and engineering, regularization and cross-validation with several models such as LASSO regression, Ridge regression, Elastic Net regression, and Polynomial regression. I also picked up important skills such as interpreting the OLS statistics table, and understanding the QQ-plot.
Medium Blog: https://medium.com/@tanpengshi/predicting-market-value-of-fifa-soccer-players-with-regression-5d79aed207d9