 What is Anumana Pramana to be precise?
 Is Hindu spiritual space immune to Muslim, Christian and Buddhist influence?
 Did Buddha ever say how to overcome shyness?
 How can I add a dynamic “active” CSS class to the navigation on any given page?
 Anime guess Riddle #4
 A Riddle of Mine
 A Geographyish Riddle
 A Christmas Tennis Tournament
 Puzzles like Sokoban?
 Can you make a Rubik's Cube pattern with a 4 color flower on every side?
 Policy gradient  and autodifferentiation (Pytorch/Tensorflow)
 Use text similarity (cosine) instead of machine learning to classify companies into industries
 Does this type of classification exist?
 Load keras model in Java
 What GPU specifications matter when training and using neural networks?
 IHS problem (spouse visa)
 Applying for a Sweden student visa while in UK as a student for an Indian citizen
 First attempt at a PCB design
 Peer to peer communication
 DC motors don't stop
Confidence intervals for GLMM: bootstrap vs likelihood profile
I've built a relatively large negative binomial GLMM (~50000 observations, 10 covariates in conditional model, 1 covariate as zeroinflation model, 3 random intercepts (650, 26, 26 levels in each group respectively), and a number of random slopes). The model was fitted with the package glmmTMB in R.
I'm trying to build confidence intervals for my covariates, and I'm trying to figure out the advantages of building them by likelihood profile vs bootstrapping. Is there a reason to use one over the other?
tl;dr parametric bootstrap intervals are slightly more reliable, but much slower to compute. I would guess that either would be adequate in your case.
Likelihood profile: Likelihood profile confidence intervals are limited by the accuracy of the asymptotic approximation that differences in deviance (2 * loglikelihood, possibly with an offset based on the saturated model) are $\chi^2$distributed. In general they should be pretty good if the minimum number of gro

tl;dr parametric bootstrap intervals are slightly more reliable, but much slower to compute. I would guess that either would be adequate in your case.
Likelihood profile: Likelihood profile confidence intervals are limited by the accuracy of the asymptotic approximation that differences in deviance (2 * loglikelihood, possibly with an offset based on the saturated model) are $\chi^2$distributed. In general they should be pretty good if the minimum number of groups in any randomeffects grouping variable is large. $n=26$ is "fairly large" (various sources quote $n=40$ or $n=50$ as "large enough not to worry about it all"; Angrist and Pischke's Mostly Harmless Econometrics, which has a Hitchhiker's Guide to the Galaxy theme, gives $n=42$). So for your case these should be reasonably good, although maybe not so good if you want to compute highprecision confidence intervals (e.g. 99.9% CIs).
Bootstrapping: parametric bootstrapping is the usual approach taken for GLMMs (nonparame
20180613 20:24:45