| The outcome of sports games quiz widespread daily life, but athletics is not only astrength factor, there is great uncertainty, this randomness makes forecasting very difficult,in this paper is not to predict the right outcome.analyzed by the history of the gamebattlefield but through the methods of probability and statistics through the establishmentof a reasonable probabilistic graphical model, to assess the value of the team players, theintent to improve the prediction accuracy of forecasting for the rest of the score to makecertaincontribution.Probability statistics method is widely used in the outcome prediction by thehistorical record of data between team training, the establishment of the predictedprobability model, the model is usually based on the current value of the team’s strength,to calculate the probability of the outcome of the game expected, and then the team’s newstrength value calculated in accordance with the strength of the model update rulesaccording to the actual results of the competition. The present method is mainly on thelevel of the team predicted that did not refine the player’s level. In fact, however, considermore reasonable, because the competition between the teams is the confrontation betweenthe players and capacity value of the players in the offensive and defensive process usuallytend to play a key role, decided the game wins from the team level negative, but the playof the players, playing time, injuries, transfer and other factors will be the strength of theteam’s cause a great impact. Established the relationship between the team and the players,from the level of the players the ability to value modeling, the use of the score, hit rate,rebounds, steals, turnovers and other data indicators to assess the ability of players toestablish a reasonable shellfish Bayesian network model, and use the EM algorithm totrain the players ability to value and validate the test data set, in order to achieve thedesired results. |
PDF Full Text Request