Abstract
In this paper, a method which is used for evaluating the performance of bio-potential surface electrode (BSE) with multi-index is presented. The Fuzzy kernel C-means (FKCM) algorithm and KF statistic are employed for classifying the BSE samples and searching an optimal classification amount respectively. Subsequently, a discriminant function is constructed by support vector machines (SVM) for recognizing the new measured samples. Experimental result shows classification correction ratios of improved FKCM algorithm are 96.3% and 85% on the IRIS and BSE dataset according a priori knowledge, furthermore, the recognition correction ratios of SVM algorithm are 96.3% and 90% on the IRIS and BSE dataset.
Original language | English |
---|---|
Pages (from-to) | 880-886 |
Number of pages | 7 |
Journal | Journal of Software |
Volume | 6 |
Issue number | 5 |
DOIs | |
Publication status | Published - 20 Jul 2011 |
Keywords
- Biopotential surface electrode
- Classification
- FKCM
- Recognition
- SVM
ASJC Scopus subject areas
- Software
- Human-Computer Interaction
- Artificial Intelligence