模型选择信息量准则AIC及其在方差分析中的应用

Model selection criterion AIC and its application in ANOVA

【目的】探讨模型选择信息量准则AIC在方差分析模型选择中的必要性和意义。【方法】简要介绍了模型选择信息量准则AIC的概念,推导了AIC在方差分析模型选择的公式,并运用AIC对水稻品种比较试验数据进行了最佳ANOVA模型选择分析。【结果】对5个水稻品种比较试验进行方差分析模型的选择是必要的;AIC准则在方差分析模型选择中是一种简单有效的方法。【结论】AIC可用于方差分析模型的选择,以改进方差分析结论的可靠性。

[Objective] The aim of this study was to conduct the necessity and significance of the model selection criterion AIC in ANOVA selection. [Method] This study introduced the conception of model selection criterion AIC briefly, deduced AIC formula in ANOVA model selection and applied AIC to select the best ANOVA model in rice variety comparison trail data. [Result] ANOVA model selection was necessary in the 5 rice varieties comparison trial data. AIC was a simple and effective method in ANOVA model selection. [Conclusion] A,IC could be used to select the best model of ANOVA and improve the credibility of the ANOVA conclusion.