Football Match Results Predicting by Machine Learning Techniques

Football is a popular worldwide sport played and loved by millions of people. And people are keen to speculate on the outcome of every football match. So far, there are several existing researches that can make good prediction for the outcomes of basketball matches or Tennis matches, but they are unable to predict the outcomes of football matches properly. As such, this paper applies the ideas of machine learning to the field of football match result prediction. We select the Premier League and La Liga data in recent 5 years as experimental samples. The samples are preprocessed and divided into training samples and test samples. Supervised learning algorithms using machine learning such as LR, GBDT, RF, etc. We learn the classifiers from the training samples, and then use the learned classifiers to classify the test samples. The experimental results show that Random Forest outperforms other models, with an accuracy rate of 66.7% on the training set and an accuracy rate of 63.8% on the test set.