Parzen Window Approximation on Riemannian Manifold

作者：

Highlights：

• High-density affinity drift is countered by local neighborhood based Parzen window.

• Variable Parzen window is able to adapt to the change in the sampling of data points.

• Affinity adjustment approximates unknown local neighborhood parameters.

• Affinity adjustment captures the difference in local distribution in a neighborhood.

摘要

•High-density affinity drift is countered by local neighborhood based Parzen window.•Variable Parzen window is able to adapt to the change in the sampling of data points.•Affinity adjustment approximates unknown local neighborhood parameters.•Affinity adjustment captures the difference in local distribution in a neighborhood.

论文关键词：Parzen window,Data affinity,Graph Laplacian regularization,Manifold regularization

论文评审过程：Received 19 March 2020, Accepted 27 September 2022, Available online 1 October 2022, Version of Record 12 October 2022.

论文官网地址：https://doi.org/10.1016/j.patcog.2022.109081