Trinh Dong @tschindou

MD, Postdoc in #biostatistics @BPH_Research @OUCRU_Vietnam @KingsCollegeLon #python #rstats #js #musicaholics, into #bayesian #brain - *Views are mine* Bordeaux, France Joined February 2019

Tweets

994
Followers

117
Following

398
Likes

295

Trinh Dong @tschindou

2 months ago

. @MicrosoftEdge pls would you mind spending some time cleaning this utterly cluttered menu? Or give me the option to? No, I definitely don't do shopping, neither do I want a scrollable main menu. Please don't ruin it!

0 0 0 29 0

Download Image

Brian Locke, MD MSCI @doc_BLocke

3 months ago

If you're going to look for subgroup effects - there are good arguments to use a Bayesian analysis. Nice editorial (👇) on a recent analysis looking for (and finding) evidence of a differential effect of neuromuscular blockade by lung stiffness in ARDS

1 3 8 628 3

Download Image

Trinh Dong @tschindou

4 months ago

My second first-author paper is humbly published. We looked at the added value of #imaging data in the prognosis of brain #tuberculosis, alone and in combination with conventional biomakers. Many thanks to people ⁦⁦@OUCRU_Programme⁩ linkedin.com/posts/trinh-do…

0 0 0 39 0

Arman Oganisian @StableMarkets

9 months ago

Check out our new paper developing Bayesian causal inference methods for recurrent event outcomes! We handle several complexities such as censoring, terminal events, and treatment timing misalignment. Best part: the Bayesian models are all implementable in Stan (@mcmc_stan).

3 21 119 10K 81

Download Image

Trinh Dong @tschindou

9 months ago

#PhDone

0 0 2 346 0

Download Image

Trinh Dong @tschindou

11 months ago

Dear Professor @f2harrell I wonder if it is sensible to bootstrap the smoothing splines construction? I got a night-and-day difference between the two measures, to the point I start doubting myself. discourse.datamethods.org/t/should-i-boo…

0 0 2 82 0

Richard Riley (R²) @Richard_D_Riley

a year ago

@joie_ensor Thanks to collaborators including @joie_ensor @GSCollins @R_L_Whittle @Kym_Snell @pauladhiman @DrXiaoLiu @Denniston_Ophth @f2harrell @laure_wynants @glen_martin1 Paper builds on our stability & multiverse work bmcmedicine.biomedcentral.com/articles/10.11… onlinelibrary.wiley.com/doi/full/10.10… & extends pmsampsize

0 8 19 3K 5

Trinh Dong @tschindou

a year ago

Why some are fine spending decades of time and millions of dollars to collect data but not spending 6 more months to plan a proper analysis is beyond me.

0 0 3 74 0

Ashley M. Gjøvik @ashleygjovik

a year ago

🧵In 2020, I nearly died from mysterious industrial chemical exposure at my home. Later I discovered my employer was dumping toxic waste into our windows from their Skunkworks semiconductor fab next-door. I tipped off EPA, who raided them in 2023. EPA just sent me the report 💀⬇️

423 11K 70K 7.6M 19K

Download Image

Alex Yanko 🇺🇦 @LeopolisDream

a year ago

Welcome the new architecture: Terminator No residuals, no dot product attention, no normalization... arxiv.org/pdf/2401.17948

12 124 785 109K 629

Download Image

ᴅʀ ᴍɪʀᴄᴇᴀ ᴢʟᴏᴛᴇᴀɴᴜ ☀️ 🌊🌴 @mzloteanu

a year ago

#statstab #111 Bayesian statistics resources [quick guide] @andrewheiss Thoughts: An excellent short guide to conducting/reporting bayesian analyses. Inc. theoretical diff. you need to be aware of, PD, ROPE, CrI, etc. #brms #bayesian #tutorial #guide evalf21.classes.andrewheiss.com/resource/bayes/

1 20 104 8K 95

Frank Harrell @f2harrell

a year ago

New blog article on the unreasonableness of the prior distribution that is effectively used in much of frequentist inference: fharrell.com/post/uprior @5_utr @PhilosopherMD1 @vandy_biostat #StatThink

6 14 79 12K 50

Gabriel Kreiman @gkreiman

a year ago

Revealing Vision-Language Integration in the Brain with Multimodal Networks. Subramaniam et al, ICML 2024. Download the article here: klab.tch.harvard.edu/publications/P…

0 22 94 13K 55

Download Image

Frank Harrell @f2harrell

a year ago

#Statistics thought of the day: In frequentist stats, the long-run operating characteristic (OC) alpha is irrelevant b/c it has nothing to do with the correctness of the decision — the OC that matters and is the heart of Bayes: hbiostat.org/bayes/bet/desi…

2 10 57 10K 53

Trinh Dong @tschindou

a year ago

I don't believe the "local storage" narrative @MSFTCopilot is trying to make for the #Recall feature. As long as no isolation measure is done, no one can prevent an outer program to from spying on your screenshots. youtu.be/uHEPBzYick0?si…