STA303H1 Lecture 10: Bayesian Inference & Mixed Model Examples, Aug 13, 2019 Summer, University of Toronto, Alex Stringer
Document Summary
Yip nexp i p n nloep observables unobservables. Or ylpoh nco021 b n 0ep 02 fut. P neoep on ticot eg pcprecision of acoke. Inla outputs posteriorsummaries basedoffof itwj1ys stew140dwj wherewj wi wj1 wj11 wm pointestimate elw ly fwj cw. lydw quantilebasedoffoftlcwj1y too. Response ofoffspring ineachbrood cgroupofkids covariateconcentrationofnitrogen a herbicide so yi ofoffspringofithhea jthbrood g l 2 3 count grouped. Yi idig poissonchi logxij xi. jp t ui p ncoep. Forexample ifbt 2 and a95 oi is a 3 thentransform pi et et and cl 3 justlikeanordinary0km a cee"i. Ecy bcoo varyi wecallvca bcosthe variancefunction it captureshowthevariancerelates to the lo bcos mean. Mi edfis a yi xip a gills hi g a strictlymonotone invertible linearpredictor g linkfunction. 121 ctet p argmaxpbcp cxtwx5 where w c i wn are finaliterationofirwls. Linksealeanapproximate95 cifor13 is bj 2stiffs wherestocks tocxtwx5f is a diagonalmatrixofweightfromthe. Naturalscale anapprox95 ci g pg 2strips g pet 25h15m. Eg if 139 2 and95 oi is 4 3 andgex logus then i.