Soil pH value grey relation estimation model based on hyper-spectral

Abstract

Purpose

The purpose of this paper is to establish the grey relational estimating model of soil pH value based on hyper-spectral data.

Design/methodology/approach

As to the uncertainty of the factors affecting the soil pH value estimation based on hyper-spectral, the grey weighted relation estimation model was set up according to the grey system theory. Then the linear regression correction model is established according to the difference and grey relation degree information between the estimated samples and their corresponding pattern. At the same time, the model was applied to Hengshan county of Shanxi province.

Findings

The results are convincing: not only that the linear regression correction model of grey relation estimating pattern of soil pH value based on hyper-spectral data is valid, but also the model’s estimating accuracy is higher, which the corrected average relative error is 0.2578 per cent, and the decision coefficient R²=0.9876.

Practical implications

The method proposed in the paper can be used at soil pH value hyper-spectral inversion and even for other similar forecast problem.

Originality/value

The paper succeeds in realising both the soil pH value hyper-spectral grey relation estimating pattern based on the grey relational theory and the correction model of the estimating pattern by using the linear regression.

Citation

Miao, C., Li, X. and Lu, J. (2018), "Soil pH value grey relation estimation model based on hyper-spectral", Grey Systems: Theory and Application, Vol. 8 No. 4, pp. 436-447. https://doi.org/10.1108/GS-05-2018-0027