Characterization of Classifier Performance on Left and Right Limb Motor Imagery using Support Vector Machine Classification of EEG signal for left and right limb movement

This work proposes an algorithm that automatically classifies electroencephalography(EEG) signal for movement of left and right hands using time domain and information theoretic features like Power Spectral Density (PSD) and Shannon Entropy with the use of support vector machine. Brain-computer interfacing (BCI) has gained momentumover the last few decades and has emerged as a promising field by providing a real time platformfor interaction between brain and automated devices which can be used for rehabilitative purposes. BCI provides considerable help in overcoming sensorimotor disabilities. The EEG recordings of left and right motorimagery are identical to the actual movement of the corresponding left and right limbs. Support Vector Machine (SVM) provides an accuracy of 91.25% thereby reaffirming its efficiency in classification of EEG signals.