| URN |
etd-0723113-091119 |
Statistics |
This thesis had been viewed 462 times. Download 10 times.
|
| Author |
Lee-peng Hui |
| Author's Email Address |
No Public. |
| Department |
Institute of Industrial Management |
| Year |
2012 |
Semester |
2 |
| Degree |
Master |
Type of Document |
Master's Thesis |
| Language |
zh-TW.Big5 Chinese |
Page Count |
44 |
| Title |
Game Forecast for the Professional Basketball League Using Neural Network |
| Keyword |
Game forecast
professional basketball league
neural network
neural network
professional basketball league
Game forecast
|
| Abstract |
The purpose of this study is to find the critical factors affecting outcomes of basketball games. The neural network model training object is based on the achievements of the Miami Heat and Boston Celtics in the 2010-2012 regular season. the results of 2012 to 2013 season and championship are used as primary module for prediction .The forecast is built by analyze and integration of the data records after each routine competition between Miami Heat and Boston Celtics to establish the critical factors via neural network technology. In respect of offense, critical factors including FG%, REB, AST ; REB,BLK and STL are significant elements to defence. High stability, and easy to converge and easy to learn are the major reasons to adopt the back-error propagation make neural network model. We use this model to predict 18 games. Hit 14 games, and the accuracy rate reach 77.78¢H.therefore, the neural network model is a practical tool to predict basketball game. |
| Advisor Committee |
Yang Tai Fu - advisor
Chen Shin Guang - co-chair
Chen Yen Nien - co-chair
|
| Files |
indicate in-campus access only |
| Date of Defense |
2013-06-25 |
Date of Submission |
2013-07-24 |
