基于氮肥运筹下水稻产量与品质协同的农艺生理指标解析

覃金华, 洪卫源, 冯向前, 李子秋, 周子榆, 王爱冬, 李瑞杰, 王丹英, 张运波, 陈松

Analysis of agronomic and physiological indicators of rice yield and grain quality under nitrogen fertilization management

QIN Jin-Hua, HONG Wei-Yuan, FENG Xiang-Qian, LI Zi-Qiu, ZHOU Zi-Yu, WANG Ai-Dong, LI Rui-Jie, WANG Dan-Ying, ZHANG Yun-Bo, CHEN Song

图6 不同回归模型对水稻产量和GQI 的预测精确度(基于动态及静态农艺指标) A、C、E分别是线性回归、支持向量回归、岭回归模型对水稻产量的预测; B、D、F分别是线性回归、支持向量回归、岭回归模型对GQI的预测。Actual GY: 真实产量; Predict-lrGY: 线性回归模型预测产量; Predict-svrGY: 支持向量回归模型预测产量; Predict-ridgeGY: 岭回归模型预测产量; ActualGQI: 真实GQI; Predict-lrGQI: 线性回归模型预测GQI; Predict-svrGQI: 支持向量回归模型预测GQI; Predict-ridgeGQI: 岭回归模型预测GQI。

Fig. 6 Prediction accuracy of different regression models for rice yield and GQI (based on dynamic and static agronomic indicators) A, C and E represent linear regression, support vector regression, and ridge regression models for predicting rice yield, respectively; B, D and F represent linear regression, support vector regression, and ridge regression models for predicting grain quality index (GQI), respectively. Actual GY: actual yield; Predict-lrGY: predicted yield using Linear Regression Model; Predict-svrGY: predicted yield using Support Vector Regression Model; Predict-ridgeGY: predicted yield using Ridge Regression Model; ActualGQI: actual GQI; Predict-lrGQI: predicted GQI using Linear Regression Model; Predict-svrGQI: predicted GQI using Support Vector Regression Model; Predict-ridgeGQI: predicted GQI using Ridge Regression Model.