基于主成分分析的NBA球员综合能力评价

Evaluation of NBA Players′ Comprehensive Ability Based on Principal Component Analysis

目前对球员综合能力的评价方法有TOPSIS法、灰色关联分析法等,但当评价指标量多时,存在计算效率低等缺陷,同时有些评价指标并不是值越大越好,且会对最优排序和最劣排序造成影响。为求取最适合的算法、作出准确的NBA球员综合能力评价,采用主成分分析法,以现役NBA联盟中538名运动员的得分、助攻、三分命中率等13项指标为实例,进行评价方法研究。分析球员多方面能力,得出球员在各项成分中得分排名并计算出各项能力最强的前十名球员,与体育界分析结果进行比较,分析球员的强项和弱项。实验结果表明,主成分分析能够高效地将数据降维,表现各项数据之间关联性,并且分析结果正确,适用于NBA球员综合能力评价。

At present,the evaluation methods of the player′s comprehensive ability mainly include Topsis method,gray correlation analysis and so on.However,the calculation efficiency is low when the evaluation index is large;indicator magnitude has an impact on optimal scheduling and worst ranking while not all large evaluation indicators equal better performance.Aiming to find the most suitable algorithm to make accurate evaluation of NBA players′comprehensive ability,we use the principal component analysis to research the evaluation method with 13 indicators involving 538 NBA league athletes′scores,assists,three point rates and other performance as examples.We analyze the players′ability in various aspects,calculate the final scores,and select the top ten players to compare with the official analysis of NBA;and then we analyze the strengths and weaknesses of the players.The experimental results show that the principal component analysis can reduce the data dimensionality,present the relationship between the performance of the data and the results are correct.Therefore,it is applicable in NBA players＇ comprehensive ability evaluation.