Implementation of Machine Learning Techniques to Predict Player Performance using Underlying Statistics

Saifi, Murtaza (2020) Implementation of Machine Learning Techniques to Predict Player Performance using Underlying Statistics. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration manual) Download (1MB) | Preview

Abstract

Fantasy sports have become a growing industry that is earning major revenue for the sports and media business. Player performances are being analysed through countless of data points to determine high quality players or match outcomes. This study focuses on analysing player performance in the English Premier League on the basis of the statistics present in their Fantasy tournament (Fantasy Premier League) with the addition of underlying statistics such as Expected Goals and Expected Assists and aims to determine how strong an attribute can underlying statistics be while predicting the player performance. Ensemble based classification modelling such as Random Forest and Extreme Gradient Boosting have been used with and without the presence of underlying statistics and a 0.19% increase in accuracy 0.2% increase in F1 score is observed with the presence of these data points. The models are validated using k-fold cross validation and a comparison analysis was conducted between 4 sampling techniques and evaluated with Accuracy, Specificity, sensitivity, Precision, recall and F1 score.

Item Type:	Thesis (Masters)
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QA Mathematics > Computer software T Technology > T Technology (General) > Information Technology > Computer software
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Dan English
Date Deposited:	11 Jun 2020 10:15
Last Modified:	11 Jun 2020 10:15
URI:	https://norma.ncirl.ie/id/eprint/4270

Actions (login required)

View Item