Sameer Deshpande
@skdeshpande91
Assistant professor in Statistics at UW–Madison. Interested in #Bayesian statistics, sports analytics, causal inference. Also cocktails and Dallas sports. #mffl
As the great American philosopher Russell Stringer Bell said, I want you to put the word out there that we back up. Stop by the SBSS+ ISBA table in the exhibition hall at #JSM2026 and find out about all the exciting Bayesian sessions taking place (and get some candy) @isba-bayesian.bsky.social
Dr J: Rich simple, vanilla extract, acidulated OJ, split base of smith&cross and planteray 3 star.
A lot of these points apply to the use of generative AI in statistics. Especially this:
@dmagazine.bsky.social 's coverage of this team over the last few years has been sensational. This article about the AD trade is right up there w/ the single best paragraph about what it means to be from Dallas (s/o @mikelikessports.bsky.social)
This partially motivated me to make cherry bounce the last few years, as I wasn’t certain that I’d get cherry heering. Happily enough, it made an even better Singapore slings & blood & sands
Messed around with this year's #BigDataBowl data. In each frame, I (i) simulated remaining path (lines) while conditioning on all prev. positions (dots) ; and (ii) computed posterior predictive prob. of completion, INT, etc. Lots left to refine (esp. conditional prior mean of DB's trajectory)
Great presentation at #CMSAC by Paul Nguyen about our recent work on modeling decathlon performance with @jsmoliga.bsky.social & Ben Lindaman
I'm guessing the next step is to assess a model based on the quality of its predictions (posterior predictive checking for Bayesian). On that view, the last paragraph from Gelman, Meng, & Stern is perhaps spiritually similar to what you're looking for
@karenhao.bsky.social ‘s “empire of ai” is brilliant and articulates so many of my thoughts & feelings about AI so much more cogently than I ever could. But I also admire the way the book distills import technical aspects into plain language.
Very belatedly going through other photos from that night. The cocktail looked even better from the top!
Finally had a chance to put the raspberry & lime cordials I made recently to good use: maiden’s blush & gimlets tonight
Its “Gibbs as a composition of lots of MH steps that always get accepted” day in my graduate Bayes class. Otherwise known as “break out all the colorful chalk day”.
The evening took a turn towards flower themed cocktails: a daisy and a water lily
Back in June, @mikelikessports.bsky.social wrote about Dallas in a way that resonated deeply w/ me. I think about and return to his words a lot. Today, he pulls no punches: the FO & ownership don't care about the city. They best keep the phrase #mffl out they mouth.
Hard to think of a better way to spend a rainy Sunday than strolling around my old college and reading @willdalrymple.bsky.social’s wonderful new book
Had a free morning in London so I paid the good reverend a visit.
This month’s been a decade with some very high highs and very very low lows. One of those highs: finally opening the cherry bounce I started back in July. The sharpness of the rye & tartness of the door county cherries have mixed much better than any of my MCMC simulations ever will.