이중 다단계 일반화 선형모형의 평균과 산포의 결합 모델링을 통한 열화시험데이터 분석

Abstract

Regression analysis was extended to include GLMs, LMMs and GLMMs as the random effect was added and the model became generalized. HGLMs have been proposed, which can expand the distribution of the random effect to a distribution other than the normal distribution (Lee and Nelder, 1996). DHGLMs is an extension of HGLMs, which also takes into account the random effect of the dispersion model (Lee and Nelder, 2006). In this study, we apply the fixed effect to the dispersion model through joint modeling (Lee and Nelder, 1991). The data used in the analysis are the degradation test data, which has the tire crack length and distance of driving as variables, and analyzed the influence of tire crack length on driving once. R package dhglm () was used for model fit of each model (Lee and Noh, 2018).