Football Match Results Based On Evidence Theory Prediction Method

As football is becoming more and more popular at present, people pay more attention to the commercial value of the football matches, but the commercial value depends on the results of the football matches ultimately. Therefore, the results forecast for the football matches has become a topic of intense debate in the sports commercialized operation up to now, and it is also an important research issue in the academic research field. In previous studies, the researchers commonly used the methods of probability theory and mathematical statistics to calculate the uncertainty of the results of the football matches under the condition that the input influencing factors are complete, so as to forecast the results of the football matches. However, this method is difficult to calculate the uncertainty of the football matches and forecast the results of the football matches when the input influencing factors are not complete. D-S theory (Dempster-Shafer theory),as a kind of inexact reasoning theory, the main characteristics of it satisfy the weaker condition than Bayesian probability theory.D-S theory has the ability to deal with unknown information and express the uncertain or unknown information. D-S theory has been widely used in medical diagnostics, target identification, military command and so on.Football is a competitive sport. The results of the football matches are unknown, but they are predictable. The results of the football forecasted by the researchers are based on the scientific analysis and the calculation of the uncertainty of the results, rather than luck. Based on the features and application scenarios of D-S theory, the method for calculating the data uncertainty can be used to forecast the results of football matches. This thesis made use of the characteristic that D-S theory can be used to calculate the data uncertainty when the priori probabilities are unknown, and proposed a method to calculate the uncertainty of the results of the football matches when the influencing factors are contradictory or the input influencing factors are not complete.Generally, the main work and contribution of this thesis are summarized as follows:1. Establishing the mathematical model for forecasting the results of the football matches based on D-S theoryIn view of the applicability of D-S theory for calculating the data uncertainty, this thesis abstracted the initialized input influencing factors into the specific uncertain data, and established the mathematical model based on D-S theory for using the evidence fusion method to calculate the uncertainty of the results of the footballs matches.2. Calculating the uncertainty of the results of the footballs matches based on the D-S theoryIn this thesis, the prediction scenarios were divided into two categories, namely there are two kinds of initialized input influencing factors, one is complete, and the other is not complete. This thesis proposed the method and the steps to calculate the uncertainty of the results of the football matches (include win, lose and draw).3. Experiment at the results ofIn this thesis, the experiment first forecasted the results of some football matches which the input influencing factors of the football matches are randomly generated, and analyzed the effectiveness of the method, and then the predicted results are contrasted with the results in the real football matches, last the efficiency and error of this method were analyzed.