Primary Large Sample Partitioning for Diagnosis and Recognition Problem Solving on the Basis of Computational Intelligence Methods
The new method of training and test sample forming from primary sample is proposed. It preserves in a generated sub-sample the most important topological properties of the original sample and did not even needs to load of the original sample into computer memory. It provides a sequential exemplar pr...
Saved in:
| Date: | 2013 |
|---|---|
| Main Author: | |
| Format: | Article |
| Language: | rus |
| Published: |
Інститут проблем реєстрації інформації НАН України
2013
|
| Subjects: | |
| Online Access: | http://drsp.ipri.kiev.ua/article/view/103417 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Journal Title: | Data Recording, Storage & Processing |
Institution
Data Recording, Storage & Processing| Summary: | The new method of training and test sample forming from primary sample is proposed. It preserves in a generated sub-sample the most important topological properties of the original sample and did not even needs to load of the original sample into computer memory. It provides a sequential exemplar processing and performs transformation of the multi-dimensional coordinate set to the one-dimensional, which is also discretized to improve the data generalization properties. This allows to significantly reduce the sample size, and to significantly decrease the requirements to computer resources. Tabl.: 1. Refs: 12 titles. |
|---|