基于Node2Vec-LGBM模型的CBA球員位置預(yù)測

打開文本圖片集

中圖分類號：TP181 文獻(xiàn)標(biāo)識碼：A 文章編號：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

隨著比賽數(shù)據(jù)的逐漸增多和分析技術(shù)的進(jìn)步，對球員位置的分析已成為運(yùn)動研究的重要方向。（剩余8578字）