Browsing by Author "Chien, Li-Chu"

Now showing 1 - 3 of 3

Multiple deletion diagnostics in beta regression models
(2013-08-01)
Chien, Li-Chu
;
統計學研究所
;
Institute of Statistics
We consider the problem of identifying multiple outliers in a general class of beta regression models proposed by Ferrari and Cribari-Neto (J Appl Stat 31:799-815, 2004). The currently available single-case deletion diagnostic measures, e.g., the standardized weighted residual (SWR), the Cook-like distance (LD), etc., often fail to identify multiple outlying observations, because they suffer from the well-known problems of masking and swamping effects. In this article, we develop group deletion diagnostic measures, such as generalized SWR, generalized LD, generalized DFFITS and generalized DFBETAS, and suggest a simple procedure for identifying multiple outliers using these. The performance of the proposed methods is investigated through simulation studies and two practical examples.
Parametric simultaneous robust inferences for regression coefficient under generalized linear models
(2014)
Chien, Li-Chu
;
Tsou, Tsung-Shan
;
統計學研究所
;
Institute of Statistics
In this article, the parametric robust regression approaches are proposed for making inferences about regression parameters in the setting of generalized linear models (GLMs). The proposed methods are able to test hypotheses on the regression coefficients in the misspecified GLMs. More specifically, it is demonstrated that with large samples, the normal and gamma regression models can be properly adjusted to become asymptotically valid for inferences about regression parameters under model misspecification. These adjusted regression models can provide the correct type I and II error probabilities and the correct coverage probability for continuous data, as long as the true underlying distributions have finite second moments.
Simultaneous estimation of the locations and effects of multiple disease loci in case-control studies
(2015-04-01)
Chien, Li-Chu
;
Chiu, Yen-Feng
;
Liang, Kung-Yee
;
Chuang, Lee-Ming
;
統計學研究所
The genetic basis of complex diseases often involves multiple causative loci. Under such a disease etiology, assuming one disease locus in linkage disequilibrium mapping is likely to induce bias and lead to efficiency loss in disease locus estimation. An approach is needed for simultaneously localizing multiple functional loci within the same region. However, due to the increasing number of parameters accompanying disease loci, these estimates can be computationally infeasible. To circumvent this problem, we propose to estimate the main and two-adjacent-locus joint effects and a nuisance parameter at the disease loci separately through a linear approximation. Estimates of the genetic effects are entered into a generalized estimating equation to estimate disease loci, and the procedure is conducted iteratively until convergence. The proposed method provides estimates and confidence intervals (CIs) for the disease loci, the genetic main effects, and the joint effects of two adjacent disease loci, with the CIs for the disease loci providing useful regions for further fine-mapping. We apply the proposed approach to a data example of case-control studies. Results of the simulations and data example suggest that the developed method performs well in terms of bias, variance, and coverage probability under scenarios with up to three disease loci.