Feature Selection of Hyperspectral Data Using an Improved Slime Mould Algorithm

Hangjian Zhou, Liancun Xiu, Yule Hu, Yingxu Xiao, Zhizhong Zheng

2022

Abstract

Hyperspectral data contains rich information but also has the problem of data redundancy, so it is necessary to extract features from the data according to the application requirements to obtain useful waveform information. Traditional hyperspectral data feature selection approaches rely on band screening and other methods, which are imprecise and inefficient. Feature selection of hyperspectral data can be viewed as an optimization process, and the Slime mould algorithm (SMA) in machine learning is an effective optimization algorithm that simulates the foraging behavior of mucilaginous bacteria. In this paper, SMA is applied to the feature selection of hyperspectral data, correlation information between the bands and the results is added to the initial sampling process of the SMA, which speeds up the convergence of SMA and reduces the error of feature selection. Based on the feature bands selected by this improved SMA, a hyperspectral soil heavy metal inversion model was constructed, and the model was evaluated using three distinct evaluation methods: root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R2). The experimental results demonstrate that the optimized model has faster convergence and less result error during the feature selection phase, and that the final inversion model is more accurate.

Paper Citation

in Harvard Style

Zhou H., Xiu L., Hu Y., Xiao Y. and Zheng Z. (2022). Feature Selection of Hyperspectral Data Using an Improved Slime Mould Algorithm. In Proceedings of the 3rd International Symposium on Automation, Information and Computing - Volume 1: ISAIC; ISBN 978-989-758-622-4, SciTePress, pages 644-650. DOI: 10.5220/0012007000003612

in Bibtex Style

@conference{isaic22,
author={Hangjian Zhou and Liancun Xiu and Yule Hu and Yingxu Xiao and Zhizhong Zheng},
title={Feature Selection of Hyperspectral Data Using an Improved Slime Mould Algorithm},
booktitle={Proceedings of the 3rd International Symposium on Automation, Information and Computing - Volume 1: ISAIC},
year={2022},
pages={644-650},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012007000003612},
isbn={978-989-758-622-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 3rd International Symposium on Automation, Information and Computing - Volume 1: ISAIC
TI - Feature Selection of Hyperspectral Data Using an Improved Slime Mould Algorithm
SN - 978-989-758-622-4
AU - Zhou H.
AU - Xiu L.
AU - Hu Y.
AU - Xiao Y.
AU - Zheng Z.
PY - 2022
SP - 644
EP - 650
DO - 10.5220/0012007000003612
PB - SciTePress