Multi-manifold Discriminant Isomap for visualization and classification

作者：

Highlights：

• Preserve the geometrical structure and enhance the discriminating capability.

• SMACOF algorithm is introduced to solve the optimization problem.

• Two metrics are designed to evaluate dimensionality reduction methods.

• Experiments show the effectiveness of the method in visualization and classification.

摘要

•Preserve the geometrical structure and enhance the discriminating capability.•SMACOF algorithm is introduced to solve the optimization problem.•Two metrics are designed to evaluate dimensionality reduction methods.•Experiments show the effectiveness of the method in visualization and classification.

论文关键词：Multi-manifold learning,Nonlinear dimensionality reduction,Visualization,Classification

论文评审过程：Received 22 June 2015, Revised 1 February 2016, Accepted 2 February 2016, Available online 11 February 2016, Version of Record 21 March 2016.

论文官网地址：https://doi.org/10.1016/j.patcog.2016.02.001