透過您的圖書館登入
IP:216.73.216.59
  • 學位論文

偏斜分佈之分層模型研究

A Study on Hierarchical Models for Skewed Responses.

指導教授 : 蘇南誠
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


我們探討來自發育神經毒的偏斜常態資料之劑量反應模型, 在分層的結構下推導出反應的非條件分配, 並且探討來自同一窩的樣本內相關性問題, 最後利用最大概似估計法去估計模型的參數, 並觀察相關係數會對這個模型的影響。

並列摘要


The distribution of the mean of a random sample drawn from a skew-normal population was derived by Su and Gupta (2015). Here, we consider a hi- erarchical structure and derive the distribution of the sample mean when the location parameter itself is a random variable with a skew-normal dis- tribution. We propose a dose-response model for the data obtained from developmental neurotoxicological experiments. To account for the critical issue of the correlation among responses from pups in the same litter, the so called intralitter correlation, a hierarchical distributional structure is used to derive the underlying unconditional distribution of responses. The maximum likelihood method is used to estimate model parameters. An example with real experimental data is used to provide further illustration.

參考文獻


Arellano-Valle, R. B., and Genton, M. G. (2005). On fundamental skew
Azzalini, A. (2005). The Skew normal Distribution and Related Mul-
developmental toxicology. Handbook of statistics, 12, 507-533.
Liang, K. Y., and Hanfelt, J. (1994). On the use of the quasi-likelihood
method in teratological experiments. Biometrics, 872-880.

延伸閱讀