Permutation importance based modified guided regularized random forest in human activity recognition with smartphone

Dipanwita Thakur, Suparna Biswas

Abstract: Highlights•A novel RF-based feature selection method, namely mGRRF, is proposed.•Permutation importance is used in the proposed feature selection method.•UCI human activity sensory data is used to test the proposed method.•Apply classification algorithms to recognize various human activities.•Analyze the performance of the proposed method using different classifiers.