论文部分内容阅读
笔者认为,到目前为止,我们尚未找到一个对任何树种、地区、年龄以及任何林分因子之间相互关系都适用的回归模型,这样,就存在着一个选择最佳模型的问题。所谓最佳模型就是那些最能反映变量之间的相关关系、预测精度较高、而且经济实用的回归方程。从林学的角度来看,最佳模型便是那些最能代表和反映林木或林分的生长规律、估测误差较小、便于应用于林业生产实际的回归方程。我们通常把相关指数最大,剩余标准差最小的回归方程式选作最佳模型。此
In the author’s opinion, so far we have not found a regression model that applies to the interrelationship of any tree species, region, age, and any stand factor, so that there is a question of choosing the best model. The so-called best model is those that best reflect the correlation between variables, high prediction accuracy, and economical and practical regression equation. From the perspective of forestry, the best models are the regression equations that best represent and reflect the growth pattern of trees or stands, with less estimation error and are easy to apply to the actual production of forestry. We usually choose the regression equation with the largest correlation index and the smallest remaining standard deviation as the best model. this