Team Performance Prediction in Cooperative Game: Using NBA as an Example
碩士 === 國立政治大學 === 資訊科學學系 === 103 === Recently much work has been done on team formation in social network mining. Little attention has been paid to the team performance prediction problem. Given the game logs along with the performance information for each team, the task of team performance predicti...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Online Access: | http://ndltd.ncl.edu.tw/handle/6q2b44 |
| Summary: | 碩士 === 國立政治大學 === 資訊科學學系 === 103 === Recently much work has been done on team formation in social network mining. Little attention has been paid to the team performance prediction problem. Given the game logs along with the performance information for each team, the task of team performance prediction is to predict the performance for a specified set of team members. A classification-based approach is proposed in this thesis. Three types of features are considered, namely the team strength feature, the team style feature and the team cooperation feature. The experiments which take the National Basketball Association as an example show that the proposed approach has good prediction accuracy and is superior to existing approach.
|
|---|