Predicting soil moisture content of tea plantation using support vector machine optimized by arithmetic optimization algorithm

Soil moisture content (SMC) is an important parameter that affects tea growth. Reasonable soil improves quality and ensures yield. Therefore, it necessary to regularly monitor the water content. However, traditional prediction algorithm has problems of low accuracy efficiency. This paper constructs evaluates performance a hybrid arithmetic optimization (AOA) support vector machine (SVM) model (AOA-SVM) for predicting SMC in plantations. Grey relation analysis (GRA) Pearson correlation are adopted select features model, then with temperature (ST), atmospheric (AT), electrical conductivity (SEC) was analyzed. The optimal penalty ( c) kernel function g) SVM determined by AOA. mean square error (MSE) coefficient determination ([Formula: see text]), absolute (MAE), (ME) were calculated evaluate model. Meanwhile, AOA-SVM compared optimized sparrow search (SSA-SVM), extreme learning (ELM), (SVM), convolutional neural networks (CNN). results showed AOA-SVM, SSA-SVM CNN better, above 93%, ELM 81.69% 89.61%, respectively. best R 2 95.03%. indicates significant higher , smaller MSE than other models, which potential implications precision agriculture.