Abstract:Effective feature dimension reduction (DR) from high-dimensional remote sensing images has been a significant challenge for remote sensing object recognition. Directly adopting vector-based DR method ignores remote sensing data's inherent tensor structure information, leading to the undersample problem (USP). In addition, the existing tensor-based DR methods either require an exponential storage space increasing with the orders of the input tensor (i.e., Tucker-form methods) or are dependent on the permutation of tensor modes limiting the discriminant capability of the DR results (i.e., tensor train (TT) form methods). To conquer these problems, unlike the existing Tucker or TT form feature representation, the novel tensor ring (TR) subspace learning theory is proposed systematically and rigorously to extend the traditional vector and tensor subspace learning to the TR subspace. Then, by embedding the Fisher criterion into TR subspace, the TR discriminant analysis (TRDA) is proposed to achieve DR for remote sensing tensors with flexible tensor rank and lower storage cost. To train TRDA under different computing resources, nonrecursive and exact TRDA training methods are presented to obtain the global suboptimal and local optimal solutions, respectively. Furthermore, to adapt to the case of multisource data and unlabeled data, the multiple TRDA (MTRDA) and semi-supervised TRDA (S-TRDA) are further proposed to refine multisource features in multiple TR subspaces and absorb useful information using adaptive scatter tensor, respectively. Using optical, hyperspectral, and SAR datasets, experimental results demonstrate that the proposed TRDA can obtain better recognition accuracy and smaller storage cost than the typical vector and tensor-based DR methods.

Sufficient Dimension Folding Via Tensor Inverse Regression

On the Structural Dimension of Sliced Inverse Regression

Slicing-free Inverse Regression in High-dimensional Sufficient Dimension Reduction

A unified generalization of inverse regression via adaptive column selection

High Dimensional Elliptical Sliced Inverse Regression in non-Gaussian Distributions

Asymptotic results for nonparametric regression estimators after sufficient dimension reduction estimation

Adjusting Inverse Regression for Predictors with Clustered Distribution

Online sparse sliced inverse regression for high-dimensional streaming data

Invertible Manifold Learning for Dimension Reduction

Tensor Ring Discriminant Analysis Used for Dimension Reduction of Remote Sensing Feature Tensor

Dimension Reduction for Fréchet Regression

Robust sufficient dimension reduction via α-distance covariance

New forest-based approaches for sufficient dimension reduction

On expectile-assisted inverse regression estimation for sufficient dimension reduction

Extracting the Optimal Dimensionality for Local Tensor Discriminant Analysis

A selective review of sufficient dimension reduction for multivariate response regression

Federated Sufficient Dimension Reduction Through High-Dimensional Sparse Sliced Inverse Regression

Sparse kernel sufficient dimension reduction

On Sliced Inverse Regression with High-Dimensional Covariates

A Principal Square Response Forward Regression Method for Dimension Reduction

On the Optimality of Sliced Inverse Regression in High Dimensions