Predicting Major League Baseball championship winners through data mining
The world of sports is highly unpredictable. Fans of any sport are interested in predicting the outcomes of sporting events. Whether it is prediction based off of experience, a gut feeling, instinct, simulation based off of video games, or simple statistical measures, many fans develop their own approach to predicting the results of games. In many situations, these methods are not reliable and
lack a fundamental basis. Even the experts are unsuccessful in most situations. In this paper we present a sports data mining approach to uncover hidden knowledge within the game of baseball. The goal is to develop a model using data mining methods that will predict American League champions, National League champions, and World Series winners at a higher success rate compared to traditional models. Our approach will analyze historical regular season data of playoff contenders by applying kernel machine learning schemes in an effort to uncover potentially useful information that helps predict future champions.
© Copyright 2016 Athens Journal of Sport. ATINER. All rights reserved.
|Subjects:||baseball sports game high performance sport competition prognosis investigation method|
|Notations:||sport games technical and natural sciences|
|Published in:||Athens Journal of Sport|