基于Node2Vec-LGBM模型的CBA球员位置预测

打开文本图片集

中图分类号：TP181 文献标识码：A 文章编号：2096-4706（2025）08-0065-06

Abstract：With the accumulation of sports data and therapid development of Artificial Intellgence technology， it is particularly important touse Big Data and Machine Learming methods tooptimize player position prediction. However， traditionalmethodsoftenignorethecomplexstructuralrelationshipsbetweenplayers，whicharecrucialforpositionprediction. Therefore，this paper proposes a player position prediction model based onNode2Vecand Light Gradient Boosting Machine （LGBM）.Through data mining andanalysis，thebasicdataofCBAplayers inthre seasons are crawled，andtheLGBMmodelis usedtopredictthepositionofplayers.Combied withhyper-parameteroptimizationandNode2Vec graphembeddngalgorithm， the accuracyof the modelitself is further improved.Theexperimentalresults show thatthe modelcan notonlyeffectively optimizetheteam'sieupandtacticalarrngements，butalsoprovide strongsupportfor thetamtoenhanceitscompetitiveess and overall performance.

Keywords： Machine Learning; Light Gradient Boosting Machine; Node2Vec; prediction model

0 引言

随着比赛数据的逐渐增多和分析技术的进步，对球员位置的分析已成为运动研究的重要方向。（剩余8578字）