🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

New paper alert! "Sensitivity analysis via % of unmeasured confounding" arxiv.org/abs/1912.02793 SA is absolutely crucial for causality: holy grail is finding one that's interpretable *&* flexible Matteo Bonvini & I propose contamination model approach, w/effect bds across % confounded

“The obligation to eliminate cannabis in countries where there was widespread traditional use is a clear example of the colonial background of the Convention, and a violation of indigenous and religious rights." narratively.com/these-enterpri…

Estimating the Prevalence of COVID-19 in the United States: Three Complementary Approaches medrxiv.org/cgi/content/sh… #medRxiv

My open letter to USMLE & NBME re: their 6/4/20 announcement about abridged #Step1 & #Step2 CK exams only at "event-based testing" sites PLUS their claim of being "committed to ensuring equity." Like, RT to endorse. #MedTwitter #MedStudentTwitter #StepUpUSMLE #StepUpNBME

Our work showing that multiple data proxies can indicate COVID-19 regional growth has been featured in NYT

Important disclaimer about an article about police-involved fatalities in PNAS that made inferences using the wrong conditional probability. The "fatal"-ly incorrect analysis was published but PNAS declined to publish this letter. Sad!

Y'all. Real American Hero Anthony Fauci drops placebo, confound, cohort, retrospective, randomized, control group, efficacy, talks smack about peer review...like, basically half a semester's worth of RM is packed into his one-minute long response.

Shocking that we are 6 months into this #COVID__19 crisis and testing wait times are still far from helpful to meaningfully contribute to curb transmission. Great collaboration with 🇺🇦 [email protected] Roy Perlis Katherine Ognyanova (@ognyanova.bsky.social) Matthew Baum and j druckman

So important in all applied ML fields! Treat each CV fold as a simulated reality, only do preprocessing and feature selection within each simulation.

Great collaboration with Gal Koplewitz, Fred Lu Leon (SK), and Caroline Buckee now publicly available medRxiv with great insights from Michael Johansson

Nintendo will use the player brands built through our own grassroot efforts to promote their new games. Then, not only do they not help us grow, they actively prevent us from growing on our own. So tired of this company. #freemelee

#FreeMelee

#FreeMeleeNintendo

Excited to share our #ICLR2021 paper w/ CS & math depts Stanford University 🎊 Evaluating the Disentanglement of Deep Generative Models through Manifold Topology! w/ Eric Zelikman Fred Lu Andrew Ng Gunnar Carlsson Stefano Ermon. Acknowledging Torbjörn Lundh Samuel Bengmark. Thread 🧵

NO FUCKING WAY I JUST SEEN THIS

The Falcon 9 second stage from the Mar 4 Starlink launch failed to make a deorbit burn and is now reentering after 22 days in orbit. Its reentry was observed from the Seattle area at about 0400 UTC Mar 26.

My advisor Edward Kennedy and I have posted an article on continuous trt effects estimation: arxiv.org/pdf/2207.11825… We study nonparam models where the dose-response curve has its own smoothness, which may differ from that of the outcome regression or cond dens of trt given X.