Abstract: Highlights•New concept of similarity; relevancy beside nearness.•Proposed a highly sparse method with good classification accuracy.•Proposed a new parameter that helps to improve the final accuracy.•Validated the proposed model on several datasets.
Loading