| Summary: | 碩士 === 國立屏東教育大學 === 應用數學系 === 101 === The purpose of this study is to predict and explore the results of Super Basketball League the ninth season with scores and game-winning. When 50% (53 games), 70% (74 games), and 90% (95 games) of the game are played, using all variables and stepwise variable selection to the linear regression and logistic regression models, re-use all of the previous game, before a games, the first two games and the first three games of the average offensive data to go into the model, after 10 games, use of models to predict match results, stepwise selection can know which variables are important explanatory variables for score or win.
When 50% of the game, best prediction results for the use of all previous games of the average overall offensive and use of all data variables predicted results of logistic regression model; when 70% of the game, best prediction results for the use of the first three games of the average overall offensive and use of all data variables predicted results of linear regression model; when 90% of the game, best prediction results for the use of the before a game of the average overall offensive and use of stepwise selection predicted results of linear regression model.
Make use of stepwise variable selection, the most repeatedly elected to the first four explanatory variables, defensive rebounds, three-point shooting, two-point shooting and offensive rebounds. At different time points and different models, stepwise variable selection out of the explanatory variables are different.
|
|---|
