美國職棒大聯盟(MLB, Major League Baseball)是全世界具有龐大關注度的運動之 一，近年來除了關注球員以及球隊的表現外，球員的薪資也是球迷討論中的焦點之一， 總會引起球迷的討論，也會開始檢視該球員的表現是否真的符合他的身價。 所以如何評估球員薪資的依據一直是很熱門的話題，最直接的依據就是球員在比賽 中的成績表現，除了球員本身在賽場上所呈現的數據表現外，許多學者也提出一些可能 會影響球員薪資的變數。目前已經有許多關於大聯盟薪資的研究，影響薪資的原因有很 多種，甚至有學者將球員分成投手與打者兩者進行分析。 因此本研究致力於將球員當年度的薪資與隔年度的薪資漲幅做區間，利用機器學習 的方法，如極限梯度提升(XGBoost)、支援向量機(SVM)與 K 鄰近法(KNN)建構分類 (Classificaition)預測模型，除了建構預測球員薪資漲幅的模型，也利用極限梯度提升去 驗證我們在本研究所新增的變數，結果顯示本研究所新增的變數可以做為預測薪資的依 據。 ;Major League Baseball is one of the most watched sports in the world. In recent years, in addition to focusing on the performance of a player and his team, a player′s salary has also been a focus of fan discussion, always generating discussion and beginning to examine whether a player′s performance really matches his worth. Therefore, how to evaluate the salary of players has always been a hot topic. The most direct basis is the performance of players in the game. In addition to the statistical performance of players on the field, many scholars have also proposed some variables that may affect the salary of players. At present, there have been many studies on the salary of major league baseball, and there are many reasons for the influence of salary. Some scholars even divide the players into pitcher and hitter for analysis. Therefore, this study focused on the players into the compensation to the annual salary increase do interval, using machine learning methods, such as limit gradient (XGBoost) and support vector machine (SVM) and K Nearest Neighbor (KNN) to do a classficiation prediction model, in addition to build models of forecasting player salary increase, also use limit gradient to validate our new variables in this research institute, the results show that the new variables can be predicted as salary in our study.
