Abstract:We present a new manifold learning algorithm called Local Orthogonality Preserving Alignment (LOPA). Our algorithm is inspired by the Local Tangent Space Alignment (LTSA) method that aims to align multiple local neighborhoods into a global coordinate system using affine transformations. However, LTSA often fails to preserve original geometric quantities such as distances and angles. Although an iterative alignment procedure for preserving orthogonality was suggested by the authors of LTSA, neither the corresponding initialization nor the experiments were given. Procrustes Subspaces Alignment (PSA) implements the orthogonality preserving idea by estimating each rotation transformation separately with simulated annealing. However, the optimization in PSA is complicated and multiple separated local rotations may produce globally contradictive results. To address these difficulties, we first use the pseudo-inverse trick of LTSA to represent each local orthogonal transformation with the unified global coordinates. Second the orthogonality constraints are relaxed to be an instance of semi-definite programming (SDP). Finally a two-step iterative procedure is employed to further reduce the errors in orthogonal constraints. Extensive experiments show that LOPA can faithfully preserve distances, angles, inner products, and neighborhoods of the original datasets. In comparison, the embedding performance of LOPA is better than that of PSA and comparable to that of state-of-the-art algorithms like MVU and MVE, while the runtime of LOPA is significantly faster than that of PSA, MVU and MVE.

A Manifold Learning Algorithm Based on Incremental Tangent Space Alignment

Incremental manifold learning via tangent space alignment

Incremental Manifold Learning Regular Optimization Algorithm on Tangent Space and Feature Space Alignment

A New Manifold Learning Algorithm Based on Incremental Spectral Decomposition.

Tangent Space Learning and Generalization

Incremental Alignment Manifold Learning

Deep Manifold Computing and Visualization Using Elastic Locally Isometric Smoothness

Deep Manifold Computing and Visualization

An Incremental Manifold Learning Algorithm Based on Iterative Decomposition

Principal Manifolds and Nonlinear Dimensionality Reduction Via Tangent Space Alignment

A feature space alignment learning algorithm

An improved local tangent space alignment method for manifold learning

Nonlinear Dimension Reduction Via Local Tangent Space Alignment

Incremental and evolutionary manifold learning: a survey

Principal manifolds and nonlinear dimension reduction via local tangent space alignment (dept of computer science, pennsylvania state univ, university park, pa)

Investigating Manifold Learning Algorithms Based on Magnification Factors and Principal Spread Directions

LTSA-LE: A Local Tangent Space Alignment Label Enhancement Algorithm

LKLR: A Local Tangent Space-Alignment Kernel Least-Squares Regression Algorithm

Manifold Learning in Local Tangent Space Via Extreme Learning Machine

Incremental Manifold Learning by Spectral Embedding Methods

Nonlinear Dimensionality Reduction by Local Orthogonality Preserving Alignment