基于GraphX传球网络的传球质量量化研究

发布时间：2018-04-21 19:35

  本文选题：大数据应用 + 传球网络　； 参考：《计算机科学》2017年12期

【摘要】：虽然大数据技术在不断成熟,但它在竞技体育领域的相关应用研究还处于探索阶段。常规篮球统计缺乏对传球数据的记录,更缺乏对传球数据的统计分析、价值挖掘及应用等方面的研究。首先,在GraphX基础上将传球数据构建成图,为传球质量的研究奠定基础;其次,提出传球质量评估方法PESV(Pass Expectation Score Value),相比于传统的助攻数与失误数的比值ATR(Assist Turnover Ratio),PESV能更全面地评价球员传球的质量;最后,介绍基于传球网络及传球质量评估方法 PESV的几种应用场景,包括传球质量对比赛结果的影响分析、基于PESV值的传球路线选择,并以华人球员林书豪为例,计算其2015-2016赛季的传球得分期望值。
[Abstract]:Although big data's technology is maturing, its application research in the field of competitive sports is still in the exploratory stage. The conventional basketball statistics lacks the record of passing data, the statistical analysis of passing data, the research of value mining and application, and so on. First of all, based on GraphX, the passing data is constructed into a map, which lays the foundation for the research of passing quality. Compared with the traditional ratio of the number of assists to the number of errors, ATR(Assist Turnover quality evaluation method, PESV(Pass Expectation Score value, can evaluate the passing quality of players more comprehensively than the traditional ratio of the number of assists to the number of errors. Finally, the quality of the pass can be evaluated more comprehensively than the ratio of the number of assists to the number of errors. This paper introduces several application scenarios based on passing network and pass quality evaluation method PESV, including the analysis of the influence of passing quality on the result of match, the choice of passing route based on PESV value, and takes Lin Shuhao, a Chinese player, as an example. Calculate its 2015-2016 season pass score expectations.
【作者单位】： 新疆财经大学统计与信息学院;新疆大学信息科学与工程学院;重庆大学计算机学院;清华大学软件学院;
【基金】：国家自然科学基金项目(61562078,61262088) 新疆维吾尔自治区自然科学基金(2016D01B014)资助
【分类号】：G841;TP311.13
，

本文编号：1783823