Randomized ICA and LDA Dimensionality Reduction Methods for Hyperspectral Image Classification

TL;DR

This paper introduces randomized ICA and LDA methods using Random Fourier features for hyperspectral image classification, improving scalability and capturing non-linear data structures more efficiently than traditional kernel methods.

Contribution

The paper proposes novel randomized ICA and LDA techniques with Random Fourier features to enhance scalability and non-linear data modeling in hyperspectral image analysis.

Findings

Outperform traditional kernel ICA and LDA in accuracy.

Reduce computational time significantly.

Handle non-linearities more effectively.

Abstract

Dimensionality reduction is an important step in processing the hyperspectral images (HSI) to overcome the curse of dimensionality problem. Linear dimensionality reduction methods such as Independent component analysis (ICA) and Linear discriminant analysis (LDA) are commonly employed to reduce the dimensionality of HSI. These methods fail to capture non-linear dependency in the HSI data, as data lies in the nonlinear manifold. To handle this, nonlinear transformation techniques based on kernel methods were introduced for dimensionality reduction of HSI. However, the kernel methods involve cubic computational complexity while computing the kernel matrix, and thus its potential cannot be explored when the number of pixels (samples) are large. In literature a fewer number of pixels are randomly selected to partial to overcome this issue, however this sub-optimal strategy might neglect important information in the HSI. In this paper, we propose randomized solutions to the ICA and LDA dimensionality reduction methods using Random Fourier features, and we label them as RFFICA and RFFLDA. Our proposed method overcomes the scalability issue and to handle the non-linearities present in the data more efficiently. Experiments conducted with two real-world hyperspectral datasets demonstrates that our proposed randomized methods outperform the conventional kernel ICA and kernel LDA in terms overall, per-class accuracies and computational time.