Two-stage Clustering for Longitudinal Data withSelf-Learned Entropy
碩士 === 國立中興大學 === 統計學研究所 === 105 === This study discusses the concept of entropy and its applications in cluster analysis. In order to calculate the entropy, we propose to introduce a ”two-stage clustering” so that a set of ”pseudo labels” can be tagged to each observation. For the first stage, the...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2017
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/48662316325315246096 |