Pattern Classification Techniques for EMG Signal Decomposition

The electromyographic (EMG) signal decomposition process is addressed by developing different pattern classification approaches. Single classi- fier and multiclassifier approaches are described for this purpose. Single classifiers include: certainty-based classifiers, classifiers based on the nearest neighbour deci- sion rule: the fuzzy k-NN classifiers, and classifiers that use a correlation measure as an estimation of the degree of similarity between a pattern and a class template: the matched template filter classifiers. Multiple classifier approaches aggregate the decision of the heterogeneous classifiers aiming to achieve better classification per- formance. Multiple classifier systems include: one-stage classifier fusion, diversity- based one-stage classifier fusion, hybrid classifier fusion, and diversity-based hybrid classifier fusion schemes.