Year of Graduation
Mathematical Modeling of Sports Competitions
Control Systems and Data Processing in Engineering
The purpose of this work is to write algorithms that determine the probability of outcomes of sports confrontations based on the input parameters. As part of the thesis students Zibrov Ivan, Polyakova Tatiana And Kashirtsev Daniel were collected data on the last 5 seasons of the sports in question using a parser written in Python 3.6.1. Goals in football and points in basketball were matched with the corresponding statistical distributions and the hypothesis of belonging to these distributions was tested. For a hierarchical set of points in a tennis match, a probabilistic model based on the Markov chain was considered. With the help of the revealed regularities, algorithms were built that simulate the probabilities of all required events. Checking the results of the algorithms revealed weaknesses that need improvement. All the defects were corrected with the help of multivariate regression, which allowed to bring the model to the required level of quality.