Ian Waudby-Smith retweeted
Very excited about this new paper by Tiger Zeng (tigerzhzeng.com/)
We study causal inference w/ high-dimensional discrete confounders
We give new bias/variance results & minimax lower bounds, which characterize fundamental limits of causal inference in high dimensions
Ian Waudby-Smith retweeted
For the Stats crowd: unlike P-values, E-values (en.wikipedia.org/wiki/E-valu…) can be combined easily under arbitrary dependence. But can we say the same for their sequential analog, e-processes?
Sadly the answer is *no*, motivating our recent work: arxiv.org/abs/2402.09698
(A thread)
At ICML at 11am Hawaiian time, I'll have a poster on differentially private confidence sets at board 801.
Come hang out if interested!
Paper: arxiv.org/abs/2202.08728
Ian Waudby-Smith retweeted
Join us in person or online for our Discussion Meeting next Tuesday, 23 May. Your comments are keenly invited on the paper ‘Estimating means of bounded random variables by betting'.
Register for the meeting at rss.org.uk/training-events/e…
Ian Waudby-Smith retweeted
Ramping up for this weekend’s #WiDSPittsburgh @CarnegieMellon events! Come join us for #datascience talks and networking with Keynote by Sheryl Skolnick, Strategic Advisor @UnitedHealthGrp @Optum! stat.cmu.edu/wids Registration is free! @WiDS_Worldwide @CMU_DietrichHSS
Ian Waudby-Smith retweeted
Shantanu Gupta, Emre Yolcu and Minji Yoon of @SCSatCMU and Ian Waudby-Smith of @CMU_DietrichHSS have been named 2023 @amazon Graduate Research Fellows.
cmu.is/fellows-23
Ian Waudby-Smith retweeted
I am on the job market for faculty and research scientist positions.
I develop statistical theory and algorithms to make data-based decisions that balance the need for robustness in high-stakes settings with strong performance in practice.
Some highlights below:
1/n
Ian Waudby-Smith retweeted
My advisor @edwardhkennedy and I have posted an article on continuous trt effects estimation:
arxiv.org/pdf/2207.11825.pdf 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.
Ian Waudby-Smith retweeted
Super excited to share new work "Comparing methods for statistical inference with model uncertainty" with @AdrianRaftery1 that just got published in PNAS (@PNASNews) ! #Statistics #DataScience #BMA (1/n)
Paper: pnas.org/doi/10.1073/pnas.21…
Ian Waudby-Smith retweeted
#ISSI This Thursday (8:30AM PT) we will have Neil Xu talking about “Post-selection inference for e-value based confidence intervals” (joint w/ Prof. Aaditya Ramdas & Prof. Ruodu Wang), followed by a discussion by Dr. Zhimei Ren @RenZhimei
selectiveinferenceseminar.co…
Ian Waudby-Smith retweeted
The biggest downfall of conformal prediction has been its assumption of i.i.d. (exchangeable) data. Here, we develop a version of conformal that works with a form of distribution shift: feedback loops between the model and the covariates.
Honored to work with @seafann on this.
pumped to share our latest work! if *any* trained model decides what data to test—for example, NNs designing novel proteins—we can quantify its uncertainty w/ finite-sample guarantees.
arxiv.org/abs/2202.03613
w/ @stats_stephen, @ml_angelopoulos, @jlistgarten, M.I. Jordan 1/5
Ian Waudby-Smith retweeted
Is it possible to efficiently identify the optimal intervention while remaining agnostic to assumptions about the causal structure?
In new work with Linbo Wang and @roydanroy, we study adapting to the presence of a d-separator using multi-armed bandits.
arxiv.org/abs/2202.05100
Ian Waudby-Smith retweeted
Huge congrats to @AmandaCoston on her @Meta PhD research fellowship!
research.facebook.com/blog/2…
Amanda's a joint student @mldcmu & @HeinzCollege @CarnegieMellon doing amazing work on fairness in ML, causal inference & more
Check out her papers & more here:
cs.cmu.edu/~acoston/
Ian Waudby-Smith retweeted
I'm excited to share this paper I have been working on with my advisor Mikael Kuusela and collaborator Pratik Patil on UQ for wide-bin unfolding. Herein, we introduce a cool decision theoretic framework and use a prior without being Bayesian! arxiv.org/abs/2111.01091
Ian Waudby-Smith retweeted
New review of incremental effects, with @bonv3, Alec McClean, & Zach Branson!
arxiv.org/abs/2110.10532
Usual effects compare all vs none exposed - but this doesn't make sense if some are destined to be exposed. Can instead ask "what'd happen if trt were slightly more likely?"
In this paper I propose simple "incremental intervention" effects for causal problems with complex longitudinal confounding and/or positivity violations:
dx.doi.org/10.1080/01621459.…
arxiv: arxiv.org/pdf/1704.00211.pdf
This is such a huge honour! Also, thanks to my fantastic coauthors/mentors @philipbstark and Aaditya Ramdas (stat.cmu.edu/~aramdas/).
Replying to @evotingcc
1) Technical Track:
@iwaudbysmith, Philip Stark & Aaditya Ramdas - RiLACS: Risk limiting audits via confidence sequences
2) Administrative Track:
Piret Ehin & Mihkel Solvak - Party cues and trust in remote internet voting: data from Estonia 2005-2019
Ian Waudby-Smith retweeted
Three presentations about #RiskLimitingAudits in Session 5 of the #EvoteID2021! The latest research on auditing complex #ElectionSystems was presented by @iwaudbysmith, @DamjanVukcevic & @MichelleBlom8, and @umbernhard.
The session was chaired by Josh Benaloh from @MSFTResearch.
Ian Waudby-Smith retweeted
After a year of work, my review paper focused on predictive uncertainty methods for AI model interpretability is ACCEPTED! We hope to provide an accessible source for those unfamiliar with #XAI #Interpretability
A #radiology-focused review of predictive uncertainty for #AI Interpretability in computer-assisted segmentation doi.org/10.1148/ryai.2021210… @proftomdoyle @DrMikeNose @MacDeptMed #InterpretableAI #ExplainableAI #ML
Ian Waudby-Smith retweeted
Another interest talk about doubly-robust confidence sequences: drive.google.com/file/d/1bAh…
Just saw an interesting Speed talk on JSM 2021 about discretization in continuous time causal models: youtube.com/watch?v=60K3kvlR…

































