Multi-manifold LLE learning in pattern recognition

作者：

Highlights：

• We introduce Multiple Manifold Locally Linear Embedding (MM-LLE) learning.

• This method learns multiple manifolds corresponding to multiple data classes.

• MM-LLE uses supervised neighborhood selection in learning multiple manifolds.

• It finds the optimum low-dimensional space by minimizing the nearness of manifolds.

• Results show that MM-LLE outperforms many well-known manifold learning algorithms.

摘要

Highlights•We introduce Multiple Manifold Locally Linear Embedding (MM-LLE) learning.•This method learns multiple manifolds corresponding to multiple data classes.•MM-LLE uses supervised neighborhood selection in learning multiple manifolds.•It finds the optimum low-dimensional space by minimizing the nearness of manifolds.•Results show that MM-LLE outperforms many well-known manifold learning algorithms.

论文关键词：Multi-manifolds,Manifold learning,Multiple classes,Near manifolds,Neighborhood selection

论文评审过程：Received 12 September 2014, Revised 28 January 2015, Accepted 3 April 2015, Available online 16 April 2015, Version of Record 16 May 2015.

论文官网地址：https://doi.org/10.1016/j.patcog.2015.04.003