-
This project was done under the course CS4131: Machine Learning.
-
This project aims to predict the future price of a football player based upon his performance and by considering the relevant features of a particular player.
-
Data has been web scraped from the soFIFA website to get the latest and updated values.
-
Footballers can be broadly classified into 4 categories: Attacker, Midfielder, Defender and Goalkeeper. This project uses only relevant features of a particular player according to the category the player belongs, to reduce the feature space.
-
RESULTS and other relevant information is attached below.
-
Different models were proposed for different categories of players, based upon the accuracy they were giving. But, overall the accuracy of every category is very good, greater than 98% every time. So, we can easily conclude that our price predictions are very highly accurate and anybody who wants this information can easily rely on the results.
