We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
基于全局敏感性分析与机器学习的冬小麦叶面积指数估算.
- Authors
郭晗; 陆洲; 徐飞飞; 罗明; 张序
- Abstract
In the estimation process of wheat leaf area index (LAI), the method of combining spectral variables with machine learning algorithms (MLs) has good performance, but due to too many input parameters will lead to data redundancy, which makes the calculation In order to improve the accuracy of LAI estimation and the computational efficiency of MLs, this study proposes a method combining global sensitivity analysis (GSA) with MLs (GSA-MLs for short). First, based on the PROSAIL simulation dataset, using GSA to quantify the effects of vegetation growth parameters on Sentinel-2 spectral variables; in addition, 4 variable screening strategies are used to rank all spectral variables, and the optimal variable is selected as the input parameter of MLs. Then, through partial least squares regression ( Three kinds of MLs, partial least square regression (PLSR), support vector machine (SVM) and random forest (RF), were used to estimate the leaf area index (LAI) of wheat. The results showed that: the red edge vegetation index It is mainly affected by the chlorophyll content, while the short-wave infrared-related vegetation index is mainly affected by the equivalent water thickness, and all spectral variables are affected by the interaction between the parameters. The 30 spectral variables screened by SLAI-SInteraction are used in the estimation of wheat LAI. has the best performance (R²=0.94, RMSE=0.38). And the running time is shortened by 54.13% during model inversion. This study proposes a method combining global sensitivity analysis with machine learning, which improves the machine learning method The method has good applicability to estimate the accuracy of LAI and the computational efficiency and mechanism in the application process.
- Subjects
LEAF area index; MACHINE learning; SUPPORT vector machines; RANDOM forest algorithms; SENSITIVITY analysis; LEAST squares; PARTIAL least squares regression
- Publication
Acta Agriculturae Zhejiangensis, 2022, Vol 34, Issue 9, p2020
- ISSN
1004-1524
- Publication type
Article
- DOI
10.3969/j.issn.1004-1524.2022.09.21