Evaluation of Separability Measures in GA-based Feature Subset Selection for Myoelectric Classification

This paper evaluates the separability measures applied on feature subset selection for myoelectric signal (MES). The separability measures which are considered to evaluation are Davies–Bouldin index (DBI), Fishers linear discriminant index (FLDI), Dunn’s index (DI) and generalized Dunn’s index (GDI). Four channel of myoelectric signal from upper limb muscles are used in this paper to classify six distinctive activities. Cascaded genetic algorithm (GA) has been adopted as the search strategy in feature subset selection. Results prove more accurate and reliable classification for the elite subset of features selected based on Davies–Bouldin index (DBI).