Abstract:Hyperspectral image classification has two key characteristics: spatial dependency and class imbalance. The synthetic oversampling methods for solving the between-class imbalance have the problem of spatial information loss to some extent. However, the feature space optimization methods for solving data within-class complex distribution problems have not been studied extensively. To remedy such an issue, we propose in this article an improved spectral–spatial feature rotation-based support vector machine (SVM) ensemble method (SSFRoS). More precisely, in the data preprocessing stage and from the perspective of sample space optimization, we devise a spectral–spatial neighborhood-based adaptive synthetic sampling algorithm (SSAS) that focuses adaptively on hard-to-learn samples, with artificial samples interpolated in the spectral–spatial neighborhood to further utilize the spatial information. In the feature rotation stage and from the perspective of feature space optimization, we innovatively discover that the class overlapping problem can be handled by combining feature extraction, to achieve the goal of reducing the difficulty of imbalanced classification. Besides, the SVM and voting procedure are used in the ensembled classification stage. We then conduct a series of experiments to compare the proposed SSFRoS with several baselines in terms of seven metrics. The results show that SSFRoS has significant advantages in dealing with imbalanced hyperspectral image classification problems. In addition, a detailed ablation study shows that the lowest class recall can be improved by 46.67% and 5.26% at Indian Pines and Pavia University, respectively, which demonstrates that using SSAS can enhance the recognition of hard-to-learn classes.

Reduced least squares one-class SVM in empirical feature space for imbalanced data

A Novel Svm Modeling Approach For Highly Imbalanced And Overlapping Classification

One Kind of Reduced Learning Algorithm for Support Vector Domain Description

Fast Incremental Learning Method for One-Class Support Vector Machine

Improved SVM algorithm for imbalanced dataset classification

Adaptive Subspace Optimization Ensemble Method for High-Dimensional Imbalanced Data Classification

One-Sided Fuzzy SVM Based on Sphere for Imbalanced Data Sets Learning

Dynamic least squares support vector machine

Asymmetric Least Squares Support Vector Machine Classifiers

Multikernel Correntropy Based Robust Least Squares One-Class Support Vector Machine

Robust least squares one-class support vector machine

Recursive support vector machines for dimensionality reduction.

An unbiased LSSVM model for classification and regression

Weighted Least Squares Twin Support Vector Machine with Fuzzy Rough Set Theory for Imbalanced Data Classification

Inter-class Sparsity Based Discriminative Least Square Regression.

Improved Sparse Least-Squares Support Vector Machine Classifiers.

A Spectral-Spatial Feature Rotation based Ensemble Method for Imbalanced Hyperspectral Image Classification

Unsupervised Clustering Based Reduced Support Vector Machines

A Pruning Strategy Based On Confidence Interval For Sparse Ls-Svm

A Non-Biased Form of Least Squares Support Vector Classifier and Its Fast Online Learning

A Fast Iterative Single Data Approach to Training Unconstrained Least Squares Support Vector Machines