Dublin Core
Title
Abstract
Over recent decades, machine learning technology has increasingly been used to predict sports performance. The sports industry generates extensive statistical data on players, teams, and seasons. Traditional prediction methods have shown limited accuracy. With data mining, sports organizations have recognized the outdated analysis in their data and are now utilizing it effectively.
The goal of this thesis is to explore accurate sports result predictions. Identifying significant features and analysing their impact on match outcomes is essential. Key variables include team statistics, player statistics, and historical data. These factors help managers and club directors forecast match winners and determine strategies. Machine learning techniques like KNN, Random Forest, logistic regression, and SVM are often applied to predict match results.
These predictions help coaches and managers assess player performance, evaluate skills, anticipate injuries, and strategize for upcoming games. Additionally, accurate predictions have significantly fuelled the sports betting industry, which is expanding rapidly thanks to the convenience of mobile and tablet devices.
This research proposes an AI-based framework to predict football match outcomes. Itexamines the effectiveness of system learning algorithms and reviews data mining techniques for predicting sports performance, highlighting their strengths and weaknesses. Despite previous research attempts, achieving high precision in game result predictions remains challenging.