当前位置: 首页 > 文章 > 预测模型精度指标的探讨 西北林学院学报 1990 (3) 34-40
Position: Home > Articles > Approach of Accuracy Criteria of Prediction Model Journal of Northwest Forestry University 1990 (3) 34-40

预测模型精度指标的探讨

作  者:
钟义山
单  位:
西北林学院林学系
关键词:
模型精度;Freese检验;平均相对误差;相对均方差;相对临界误差
摘  要:
本文对林业实践中常用的检验预测模型精度的指标进行了讨论,并从Freese提出的假设检验结构的方法中导出一个有明确统计意义的相对临界误差的指标。该指标可以检验模型是否满足精度要求,也可以作为比较两个模型优劣的一个指标。文中给出了木材生产中预测劳动生产率和成本的数量化模型精度检验的例子。
译  名:
Approach of Accuracy Criteria of Prediction Model
作  者:
Zhong Yishan (Dept. of Forestry)
关键词:
model accuracy, Freese test,relative mean error, relative mean square deviation, relative critical error
摘  要:
Presents a discussion on usual checking criteria of prediction model accuracy in forest practice and derives an index of relative critical error that has definite statistical sense by Freese's hypothetical test construction method. This index may be used to check whether a model meets accuracy requirements, and also to compare fitness between two models. An example of checking accuracy of quantizing model that predicts labour productivity and production cost in wood industry is given.

相似文章

计量
文章访问数: 11
HTML全文浏览量: 0
PDF下载量: 0

所属期刊

推荐期刊