An Applications of Data Mining Techniques in Talent Identification of Football Players

Please enable javascript in your browser.

Safari Majd, Mohammad | 2024

0 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 57468 (01)
University: Sharif University of Technology
Department: Industrial Engineering
Advisor(s): Akhavan Niaki, Taghi
Abstract:
Data mining in sports, particularly football, has grown significantly in recent years. Football clubs hire data scientists to conduct their transfers based on classification algorithms. These algorithms are employed so that club managers can narrow the numerous options available in the market to a limited number that fits the club's conditions. We understand that financial resources are limited; therefore, a club cannot acquire all the best players simultaneously. This creates a need for financial resource management and a thorough evaluation of a player's transfer. Using datasets from FBref and Transfermarkt, and with the help of machine learning methods, we have attempted to predict a player's buy status in various positions using eight different machine learning algorithms. The calculations indicate that the results from the neural network method show at least 92% accuracy across all positions
Keywords:
Machine Learning ; Data Mining ; Football ; Classification ; Talent Identification

No TOC