Optimum Feature Selection for the Analysis of Effects of Electroencephalography on Applied Pressure Physiotherapy

Applied Pressure Physiotherapy (APP) approaches have been widely acknowledged as helpful in treating a wide range of musculoskeletal and neurological disorders. However, the impact of these methods on mental performance has only been partially explored. The research being conducted will provide an exploratory examination of how APP affects EEG spectral patterns, which may help researchers better understand the underlying neurophysiological mechanisms that contribute to the therapeutic benefits of such therapies. Improved classifier performance is possible thanks to feature selection’s potential to lessen the impact of noise in the data, overtraining, storage requirements, and computing complexity. These objectives have been met through the application of various feature selection techniques. A novel hybrid feature selection strategy is proposed in this research. Performance is measured using a support vector machine classifier; features are extracted using a Filter-Bank similar spatial resolution pattern; and an ideal feature subset is found and generated using the grey wolf optimization algorithm. The preliminary findings demonstrate that the suggested methods outcompete the cutting-edge methods even with limited training data.