Study Of EMG Feature Selection For Hand Motions Classification
In recent days, electromyography (EMG) pattern recognition has becoming one of the major interests in rehabilitation area. However, EMG feature set normally consists of relevant, redundant and irrelevant features. To achieve high classification performance, the selection of potential features is c...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | en |
| Published: |
Penerbit Universiti, UTeM
2019
|
| Online Access: | http://eprints.utem.edu.my/id/eprint/24153/2/5094-14025-1-PB.PDF http://eprints.utem.edu.my/id/eprint/24153/ https://journal.utem.edu.my/index.php/ijhati/article/view/5094/3659 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In recent days, electromyography (EMG) pattern recognition has becoming one of the major interests in rehabilitation area. However, EMG feature set normally consists of relevant, redundant and irrelevant features. To achieve high classification performance, the selection of
potential features is critically important. Thus, this paper employs two recent feature selection methods namely competitive binary gray wolf optimizer (CBGWO) and modified binary tree growth algorithm (MBTGA) to evaluate the most informative EMG feature subset for efficient
classification. The experimental results show that CBGWO and MBTGA are not only improves the classification performance, but also reduces the number of features. |
|---|