Applying Bayesian Network to Predict Student Performance in Question-Answer Situation
碩士 === 元智大學 === 資訊工程學系 === 101 === This research developed two mechanisms: applying Bayesian Network to predict student performance in Peer Instruction question-answering situation, and assigning heterogonous neighbors based on predicted performance. The research built Bayesian Network by presenting...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Online Access: | http://ndltd.ncl.edu.tw/handle/42955161323658652500 |
| Summary: | 碩士 === 元智大學 === 資訊工程學系 === 101 === This research developed two mechanisms: applying Bayesian Network to predict student performance in Peer Instruction question-answering situation, and assigning heterogonous neighbors based on predicted performance. The research built Bayesian Network by presenting questions by nodes and by connecting nodes by links based on the associations among questions and involved knowledge. The previous student performances in question-answer situation were collected and were applied to build the conditional probability table among the nodes of the Bayesian Network. Student existed performance were inputed to the Bayesian Network to predict student performances of the next week. The system assigned heterogonous neighbors for peer instruction based on the predicted performance. The predicted performances were compared to actual performances to evaluation the prediction accuracy of the Bayesian Network. The results of two experiments showed that the prediction precisions were about 62% and 58%. The results also revealed that the heterogonous neighbor assignments did not statistically improve the student performance in Peer Instruction.
|
|---|