节点文献
基于3种回归模型的红枣产量动态预报
Dynamic Prediction of Red Date Yield Based on 3 Kinds of Regression Models
【摘要】 根据宁夏中卫市1981~2005年红枣产量数据与相应年份的气象要素数据,分别运用进入回归模型、逐步回归模型、向后回归模型分析红枣返青期、开花期、果实转色期气象因子与产量的关系,并对模型计算结果进行产量丰歉等级划分,通过数学期望判定模型优劣度。结果表明,返青期采用逐步回归模型、开花期采用向后回归模型、果实转色期采用进入回归模型预报当年产量丰歉程度准确率较高。
【Abstract】 Based on yield data of red date and corresponding meteorological element data of Zhongwei City in Ningxia in 1981-2005,the relationships between meteorological factors and yield in return-to-green stage,flowering stage and fruit color-changing stage of red date were analyzed by applying entering regression model,stepwise regression model and backward regression model.And the calculation results of models classified on the basis of poor and rich harvest grade of yield and the excellence degree of model was assessed through mathematical expectation.The results showed that the accurate rate of predicting poor and rich harvest degree of yield in current year by applying stepwise regression model in return-to-green stage,backward regression model in flowering stage and entering regression model in fruit color-changing stage was higher.
【Key words】 Red date; Yield prediction; Regression model; Mathematical expectation;
- 【文献出处】 安徽农业科学 ,Journal of Anhui Agricultural Sciences , 编辑部邮箱 ,2010年34期
- 【分类号】S665.1
- 【被引频次】6
- 【下载频次】151