GDP 可以預測奧運表現嗎?

Abstract

中文摘要

本研究主要針對2000-2016年間（共5屆奧運），探討各國GDP對其夏季奧運會表現的影響。文中透由文獻分析法將重要的總體經濟變數，特別是影響選手在奧運會上的表現，納入迴歸模型的設定中，並提出了一個新的變數——RT5，上一屆奧運排名前五的國家。另外，我們同時使用了LASSO算法進行變數選取，其中排除了共產國優勢這個變數。最後，我們對2016年奧運前10名的國家，在獎牌指數（MI）上進行預測；同時我們也預測了各國在2016年奧運所獲得的金牌數。文中發現（最適）模型所產生出的預測金牌數，在絕大多數情況下，皆能夠捕捉到各國所獲得的實際金牌數。

Abstract This study focuses on the impact of GDP on the performance of the Summer Olympics between 2000 and 2016 (five Olympic Games). Through the literature analysis method, the important economic variables which can precisely predict the Olympic performance of a country, are included into our regression model, and a new variable is proposed: RT5, the top five countries in the previous Olympic Games. In addition, we also use the LASSO algorithm for variable selection, found that “the advantages of the communist country” is no more significant. Finally, we forecast the medal index (MI) for the top 10 countries in the 2016 Olympics. Meanwhile, we also forecast the number of gold medals each country won in the 2016 Olympics. The number of predicted gold medals which generated by our (optimal) model can capture the actual number of gold medals in most cases.