基于机器学习的足球比赛数据特征研究

Research on Data Features of Football Matches Based on Machine Learning

摘要: 为研究足球比赛数据特征，本文通过机器学习对足球赛场数据进行功能性分析，特别是通过社会网络分析，量化各球员在每场球赛的表现以及团队合作的效果。本文先考虑个人对团队的贡献，对13个子指标归纳为四个大指标，综合优序图法和独立性权重法构建网络球员个人贡献模型并提出建议。为分析团队合作的效果，本文选取15个相关指标进行筛选并进行回归分析，最终确定影响比赛结果的6个显著性因素，并给出结论和建议。

Abstract: In order to study the characteristics of football game data, this paper uses machine learning to perform functional analysis of football matches data, especially through social network analysis, to quantify the performance of each player in each match and the effect of teamwork. This paper first considers the individual’s contribution to the team, summarizes the 13 sub-indices into four major indicators, integrates the optimal sequence diagram method and the independence weight method to construct the individual contribution model of the network player and makes recommendations. In order to analyze the effect of teamwork, this paper selects 15 relevant indicators for screening and regression analysis, and finally determines 6 significant factors that affect the results of the matches, and gives conclusions and suggestions.