Dynamic selection of classifiers ensemble applied to the recognition of EMG signal for the control of bioprosthetic hand

In the paper the problem of EMG-based recognition of user intent for the control of bio-prosthetic hand is addressed. The multiple classifier systems (MCS) with dynamic ensemble selection (DES) strategy based on the original concept of competence measure are applied. In the proposed method first a probabilistic reference classifier (RRC) is constructed which - on average - acts like the classifier evaluated. Next, the competence of the classifier is calculated as the probability of correct classification of the respective RRC. The performace of two MCSs with proposed competence functions were experimetally compared against four benchmark MCSs using real data concerning the recognition of seven types of grasping movements. The systems developed achieved the highest classification accuracies demonstrating the potential of DES systems with competence mesure for recognition of EMG signals.

[1]  David G. Stork,et al.  Pattern Classification , 1973 .

[2]  Marek Kurzynski,et al.  A probabilistic model of classifier competence for dynamic ensemble selection , 2011, Pattern Recognit..

[3]  Harry J. Griffiths.  Encyclopedia of Medical Devices and Instrumentation , 1989 .

[4]  Subhash C. Bagui,et al.  Combining Pattern Classifiers: Methods and Algorithms , 2005, Technometrics.

[5]  J. Berger.  Statistical Decision Theory and Bayesian Analysis , 1988 .

[6]  Thaweesak Yingthawornsuk,et al.  Assessment of vocal correlates of clinical depression in female subjects with probabilistic mixture modeling of speech cepstrum , 2011, 2011 11th International Conference on Control, Automation and Systems.

[7]  Ian T. Jolliffe,et al.  Principal Component Analysis , 2002, International Encyclopedia of Statistical Science.

[8]  Marek Kurzynski,et al.  Control of Artificial Hand via Recognition of EMG Signals , 2004, ISBMDA.

[9]  Adel Al-Jumaily,et al.  Fuzzy wavelet packet based feature extraction method for multifunction myoelectric control , 2007 .

[10]  Thomas G. Dietterich.  Approximate Statistical Tests for Comparing Supervised Classification Learning Algorithms , 1998, Neural Computation.

[11]  Fabio Roli,et al.  Dynamic classifier selection based on multiple classifier behaviour , 2001, Pattern Recognit..

[12]  Marek Kurzynski,et al.  Human-machine interface in bioprosthesis control using EMG signal classification , 2010, Expert Syst. J. Knowl. Eng..

[13]  Robert Sabourin,et al.  From dynamic classifier selection to dynamic ensemble selection , 2008, Pattern Recognit..

[14]  Robert P.W. Duin,et al.  PRTools3: A Matlab Toolbox for Pattern Recognition , 2000 .

[15]  Junuk Chu,et al.  A Real-Time EMG Pattern Recognition System Based on Linear-Nonlinear Feature Projection for a Multifunction Myoelectric Hand , 2006, IEEE Transactions on Biomedical Engineering.

[16]  Paul C. Smits,et al.  Multiple classifier systems for supervised remote sensing image classification based on dynamic classifier selection , 2002, IEEE Trans. Geosci. Remote. Sens..

[17]  Marek Kurzynski,et al.  A Measure of Competence Based on Randomized Reference Classifier for Dynamic Ensemble Selection , 2010, 2010 20th International Conference on Pattern Recognition.

[18]  Joshua C. Kline,et al.  Decomposition of surface EMG signals. , 2006, Journal of neurophysiology.

[19]  Shigeo Abe DrEng.  Pattern Classification , 2001, Springer London.

[20]  Marek Kurzynski,et al.  Control of Dexterous Hand Via Recognition of EMG Signals Using Combination of Decision-Tree and Sequential Classifier , 2008, Computer Recognition Systems 2.