基于Node2Vec-LGBM模型的CBA球员位置预测  

CBA Player Position Prediction Based on Node2Vec-LGBM Model

作　　者：何家丽 杨军 HE Jiali;YANG Jun(School of Computer and Information Engineering,Shanghai Polytechnic University,Shanghai 201209,China)

机构地区：[1]上海第二工业大学计算机与信息工程学院,上海201209

出　　处：《现代信息科技》2025年第8期65-70,共6页Modern Information Technology

摘　　要：随着体育数据的积累和人工智能技术的快速发展,利用大数据和机器学习方法优化球员位置预测变得尤为重要。然而,传统方法往往忽略了球员之间复杂的结构关系,而这些关系对位置预测至关重要。因此,文章提出一种基于Node2Vec和轻量级梯度提升机(LGBM)的球员位置预测模型。通过数据挖掘和分析,爬取了CBA球员3个赛季的球员基础数据,并利用LGBM模型进行球员位置预测,结合超参数调优以及Node2Vec图嵌入算法,进一步提高模型本身的准确率。实验结果表明,该模型不仅能有效优化球队的阵容和战术安排,还能为球队提升竞争力和整体表现提供有力支持。With the accumulation of sports data and the rapid development of Artificial Intelligence technology,it is particularly important to use Big Data and Machine Learning methods to optimize player position prediction.However,traditional methods often ignore the complex structural relationships between players,which are crucial for position prediction.Therefore,this paper proposes a player position prediction model based on Node2Vec and Light Gradient Boosting Machine(LGBM).Through data mining and analysis,the basic data of CBA players in three seasons are crawled,and the LGBM model is used to predict the position of players.Combined with hyper-parameter optimization and Node2Vec graph embedding algorithm,the accuracy of the model itself is further improved.The experimental results show that the model can not only effectively optimize the team's lineup and tactical arrangements,but also provide strong support for the team to enhance its competitiveness and overall performance.

关 键 词：机器学习 轻量级梯度提升机 Node2Vec 预测模型 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]