An integrated system for semi-automated linear correlation discovery in databases.
Measures of association (e.g. x2i ^2) are useful in facilitating knowledge discovery in databases. However, to automate the measurement of associations in data sets, several problems must be resolved. These problems include: 1) selecting appropriate association measurement functions, 2) reconciling...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Theses and Dissertations |
| Language: | English |
| Published: |
2008
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/7198 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | Measures of association (e.g. x2i ^2) are useful in facilitating knowledge discovery in databases. However, to automate the measurement of associations in data sets, several problems must be resolved. These problems include: 1) selecting appropriate association measurement functions, 2) reconciling the results obtained from heterogeneous association measurement functions, 3) ensuring that threshold values for the measures of association are appropriate for discovering interesting associations, 4) validating discovered associations, and 5) integrating the statistics packages, DBMSs and data mining tools. |
|---|