1. Home page
  2. Search

Sports Analytics Performance Prediction

This item is provided by the institution :
International Hellenic University   

Repository :
Institutional Repository of the International Hellenic University   

see the original item page
in the repository's web site and access all digital files if the item*



Sports Analytics Performance Prediction (EN)
Giannakoulas, Nikolaos (EN)
Tjortjis, Christos (EL)
Koukaras, Paraskevas (EN)
Berberidis, Christos (EN)
masterThesis
2023-04-05
2023-02-09
2023-04-05T08:45:57Z
This dissertation was written as part of the MSc in Data Science at the International Hellenic University. Sports Analytics is a rapidly growing field. It is experiencing great development, and its applications are very useful to sports clubs. These clubs can collect information about the players and the game generally and then extract important insights that will help them improve. There is a wide variety of daily football statistics. Thousands of football games happen every week, resulting in the production of numerous statistics. Sports analytics’ responsibility is to collect those statistics, analyze them and then provide conclusions to football clubs. The goal of this dissertation is to predict the performance of a player in the field of football. The purpose of the dissertations is to predict, as accurately as possible the number of goals a football player will achieve next season based on his previous years’ performances. Football players’ data were collected from valid online sources and then analyzed [1]. After that, feature engineering was implemented to transform the collected data into the desired form. Then, three machine learning algorithms were used for predictions. This dissertation is separated into two parts. The first part is theoretical, and it includes previous works on the field of performance prediction, not only in football but in other sports too, like basketball, volleyball and tennis. For the second part, the practical one, we paid attention to the conducted experiments. The results were examined thoroughly and compared to understand them and determine which model had better performance. For this purpose, python was used for coding, specifically PyCharm. (EL)
Machine learning (EL)
Performance prediction (EL)
Sports analytics (EL)
English
School of Science and Technology, MSc in Data Science
IHU (EN)
Default License
International Hellenic University
Institutional Repository of the International Hellenic University
University Center of International Programmes of Studies (UCIPS)
Master Theses



*Institutions are responsible for keeping their URLs functional (digital file, item page in repository site)