This paper presents the design and results of the submission from the researchers from Universidade Federal Fluminense (UFF) to the Machine Learning Soccer Prediction Challenge, organized by the Springer Machine Learning Journal. The steps of parsing of data, building of the model and choice of a Decision Tree for predicting the results from soccer matches are explained and discussed. This study also analyses the variation of the win percentage of home teams in different countries. Some possibilities to explain these variations, such as country size and training infrastructure are raised. The intention is to provide researchers in the field with insights for additional data that may help the prediction of games in regions with diverse patterns.
Paper
Full text
Decision Trees for the Prediction of Outcome of Soccer Games - Historical Data Analysis
Semantic Scholar · Computer Science · 2020
Abstract
This paper presents the design and results of the submission from the researchers from Universidade Federal Fluminense (UFF) to the Machine Learning Soccer Prediction Challenge, organized by the Springer Machine Learning Journal. The steps of parsing of data, building of the model and choice of a Decision Tree for predicting the results from soccer matches are explained and discussed. This study also analyses the variation of the win percentage of home teams in different countries. Some possibilities to explain these variations, such as country size and training infrastructure are raised. The intention is to provide researchers in the field with insights for additional data that may help the prediction of games in regions with diverse patterns.