Spatial Objects Matching and Transition Analysis Applied in Data Mining
| Autor: | Yang Meng-Han, 楊孟翰 |
|---|---|
| Rok vydání: | 2001 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 89 This thesis proposed an approach to apply the techniques of spatial objects maching and transition analysis in data mining. The data item with numerous attributes in database can be viewed as a object in multi-dimensional space, if we treat the attribute as a spatial dimension. Therefore, we can transform all data items in database to objects in a spatial space. With the technique of spatial object matching, we can find the similarity between different data clusters scattering over distinct time regions. And with the technique of spatial object transition analysis, we can discover the transformations of two similar data clusters in the viewpoint of multiple data attributes. That is, we can find out the patterns of data change in distinct data attributes over diverse time regions. The abstract description outputted by the proposed data mining process provides important clues for understanding the trend of an evolution process or the effects of various influential factors. So, we can mine some useful information and knowledge from the changes of data. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|