The Application of Decision Tree in the Prediction of Winning Team

Decision tree algorithm is mainly used for analyzing data and forecasting.The algorithm is easy to understand and implement, and it is easy to evaluate the model through static testing.In this paper, the decision tree algorithm is applied to predict the result of Chinese Football Association Super League(CFASL). In this paper, the data of CFASL 2015, 2016 and 2017 seasons are collected. And "the result of each team's last match", "compare of teams' ranking in last season", "the result of last time the two teams played",etc. are selected as features to predict the winner of a CFASL match. The experiments found that "two teams' performance in their recent matches", "whether team A has defeated team B as an away team" and "whether team B has defeated team A as an away team" played a key role in improving the accuracy of predicting.