Application of Symbolic Data Analysis to the Prediction of the FIFA World Cup Score

• Main Authors: CHEN, YEN-CHEN, 陳彥辰
• Other Authors: WU, HAN-MING
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/jr8a28

碩士 === 國立臺北大學 === 統計學系 === 107 === The sport data consists of the teams’ and its players’ records for games played. The collection and researches of such data have a certain reference value to the professional arena and the development of related industries. In this study, we aim to apply the method of the symbolic data analysis to predict the team’s possible score in the next competition based on their historical world cup scores. Conventionally, the sport data is collected and analyzed based on the units of the basic information of players and their performance. Instead of using the individuals as units, this study regards the teams as the units of analysis by aggregating the players’ values within the same team for some variables. For example, the age is expressed in an interval-valued variable, the nationality is represented by a barchart-valued variable, and/or the scores are represented by a histogram-valued variable. The symbolic regression analysis is then applied to predict the team’s scoring performance. The results are compared with those obtained by the traditional analysis methods.