Molas, M.

model likelihood e ffects distribution parameter approach h-likelihood study estimate estimation component poisson patient analysis value variance matrix dispersion hurdle model function score effect chapter structure method mixture covariate hurdle section correlation table stroke proteinuria response h-likelihood approach poisson distribution weight level resolution dispersion parameters outcome -13 gaussian file likelihood center nelder result algorithm procedure point mmp -13 poisson model hypertension intercept op model preeclampsia gamma number count respect treatment subject vector error statistic periodontiti e ffects models inference poisson part measurement design measure <0.0001 example women expression class overdispersion pressure postpartum