Nonlinear non-extensive approach for identification of structured information

TL;DR

This paper introduces a nonlinear non-extensive method for identifying structured information in complex data, especially when linear techniques fail due to overlapping subspaces, improving discrimination in challenging scenarios.

Contribution

The paper proposes a novel nonlinear non-extensive approach that effectively separates structured information in cases where linear methods are inadequate.

Findings

Nonlinear approach outperforms linear methods in overlapping subspace scenarios.

Method effectively discriminates structured information in complex data.

Applicable to diverse phenomena with intertwined structures.

Abstract

The problem of separating structured information representing phenomena of differing natures is considered. A structure is assumed to be independent of the others if can be represented in a complementary subspace. When the concomitant subspaces are well separated the problem is readily solvable by a linear technique. Otherwise, the linear approach fails to correctly discriminate the required information. Hence, a non extensive approach is proposed. The resulting nonlinear technique is shown to be suitable for dealing with cases that cannot be tackled by the linear one.