00:29:42 Brendan Lam: Hes got a point
00:31:21 Will Parbury: https://en.wikipedia.org/wiki/Shrinkage_(statistics)
00:31:56 Brendan Lam: http://haines-lab.com/post/on-the-equivalency-between-the-lasso-ridge-regression-and-specific-bayesian-priors/
00:32:03 Brendan Lam: “On the equivalency between frequentist Ridge (and LASSO) regression and hierarchial Bayesian regression"
00:33:45 defuneste: https://docs.google.com/spreadsheets/d/18IDSOU2bfkD55kOB18qCB7Idbpiyp4_9qeWjkvE-Syc/edit#gid=0
00:37:51 Erik Aa: https://www.youtube.com/watch?v=SocRgsf202M
00:38:02 Lisa Lau: “hierarchical Bayesian models actually contain frequentist ridge and LASSO regression as a special case—namely, we can choose a prior distribution across the β weights that gives us a solution that is equivalent to that of the frequentist ridge or LASSO methods! Not only that, but Bayesian regression gives us a full posterior distribution for each parameter, thus circumventing problems with frequentist regularization that require the use of bootstrapping to estimate confidence intervals.” 😯
00:38:17 Erik Aa: wow
00:39:20 Brendan Lam: Good luck!
15.9.2 Cohort 2
15.9.3 Cohort 4
Meeting chat log
00:01:49 Federica Gazzelloni: Hello
00:02:03 Derek Sollberger (he/him): good morning
00:02:19 Derek Sollberger (he/him): https://dsollbergerquarto.netlify.app/posts/bayesrules/14_naive_bayes_classification
00:02:48 Federica Gazzelloni: Most probably is that the package needs to be checked in the DESCRIPTION file
00:03:16 Federica Gazzelloni: Thanks for sharing!
00:03:45 Derek Sollberger (he/him): rstan is not compatible with the new 4.3.0 (as said in the rstan GitHub page). I don't know how to use the compatible version of rstan with the "build book" process
00:04:09 Federica Gazzelloni: Reacted to "rstan is not compati..." with 🧐
00:04:39 Derek Sollberger (he/him): start
00:56:08 Derek Sollberger (he/him): end