Yea, probably not an actual problem. Plus, I'd bet that almost everyone who liked this tweet (saving those that have a good deal of teaching experience), are no longer qualified to judge pedagogical efficacy for newbs. And the classic worked!https://twitter.com/generativist/status/1126477850943508480 …
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
IIRC -- (and I doubt I do) -- that was the first time I went, "oh, wait, this is cool!"
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I'd read it (before or after you pressed <publish />)!
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
You tell me Mr GoC. (Congrats, again!) But yea, I don’t think it’s really a problem at entry; it only seems trite after you grok it.
-
i don't think it would be too much to start with estimating a location parameter using a t3 and data with an outlier... stage 2 in the tutorial would be to drag the outlier around and see how its effect on the posterior depends on its distance from the rest of the points
End of conversation
New conversation -
-
-
I guess I should switch to infinite SD (0 1) truncated normal then to break from the crowd :) But in all seriousness the initial tweet hit hard. Most of my scipy tutorial is beta binom or linreg given i only have 30 minutes to showcase bayes rule in an applied way
-
Similar to a tweet above I spent decades being taught frequentist methods in numerous courses over hundreds of hours, and I only stumbled upon bayes rule way too late in life.
- 2 more replies
New conversation -
-
-
After you’ve read four medium posts in Bayesian learning you get to refer to your bad opinions as informed _priors_
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
The trouble with not including trendiness as a criterion is the ongoing need to re-explain and justify basic concepts. It's not that FP or Bayesian analysis are fundamentally harder, they're just REALLY different than methods most people know, and many assume different==wrong
-
I also think, that like learning a language, Bayesian methods would have been easier to use later in my career if I wouldn't have had an exclusively undergraduate and graduate school. It was initially a big paradigm shift! (and rare chance to meaningful use that phrase).
- 1 more reply
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.