JungHo Lee

Statistics PhD Student @ CMU interested in causal inference, machine learning, nonparametric statistics, social science

[Emily Byun @yewonbyun_]

💡Can we trust synthetic data for statistical inference?

We show that synthetic data (e.g. LLM simulations) can significantly improve the performance of inference tasks. The key intuition lies in the interactions between the moments of synthetic data and those of real data

[Stat.ML Papers @StatMLPapers]

ML-assisted Randomization Tests for Detecting Treatment Effects in A/B Experiments https://t.co/KQ0HPMNJPX

[Econometrics Papers @eBlogs]

Conditional Influence Functions. https://t.co/8AYWMttp7h

[John A. List @Econ_4_Everyone]

Lecturing on experimental economics this morning and a student asked: once I have collected data, what is the best way to enhance power?

My answer: use ML techniques rather than a linear regression approach to model covariates. In the attached study, we show that using ML techniques instead of linear regression adjustment can allow researchers to attain similar levels of statistical power with 4-8 percent fewer observations!
Code is available too.

https://t.co/I5gsauWeWx

[Andy Kim @AndyKimNJ]

We just won our Senate race! As a son of immigrants, a public school kid, I never could’ve imagined I’d get to serve as a US Senator. I’m deeply humbled and grateful to NJ and for everyone who got us here. I promise I’ll serve with honor and integrity as a public servant for all.

[The Nobel Prize @NobelPrize]

한 강 Han Kang – awarded the 2024 #NobelPrize in Literature – was born in 1970 in the South Korean city of Gwangju before, at the age of nine, moving with her family to Seoul. She comes from a literary background, her father being a reputed novelist. Alongside her writing, she has also devoted herself to art and music, which is reflected throughout her entire literary production.

[The Nobel Prize @NobelPrize]

In her oeuvre, 2024 literature laureate Han Kang confronts historical traumas and invisible sets of rules and, in each of her works, exposes the fragility of human life. She has a unique awareness of the connections between body and soul, the living and the dead, and in her poetic and experimental style has become an innovator in contemporary prose.

Learn more about this year’s #NobelPrize in Literature: https://t.co/qTBSeVQHhJ

[JungHo Lee @leejunghoj]

Very interesting paper:

https://t.co/UbHiUvJp8K

[Edward Kennedy @edwardhkennedy]

there are surprisingly many open problems when it comes to theory/methods in causal inference

check out this talk by Siva Balakrishnan for an excellent & comprehensive summary of the state of the art

https://t.co/BixhF4jImP

https://t.co/pVIslnRatf

[Sean J. Taylor @seanjtaylor]

These recent slides from Susan Athey and Guido Imbens at NBER are a great recent review of the most valuable data science methods I'm aware of. They cover tons of ground with lots of pointers.

https://t.co/BEjcLLi2vq

[Edward Kennedy @edwardhkennedy]

This paper by @ramyavinayak is one of my favorites of the past few years. Interesting & exciting problem, creative analysis, clearly written, etc etc.

https://t.co/3iEy7I6P2s

First heard about this problem from Greg Valiant, in a talk on this paper:

https://t.co/WRoflLCQOt

[Panos Toulis @PanosToulis]

Looking forward to the "Design and Analysis of Networked Experiments" workshop this summer in London!
I will join to discuss experimenting on interfirm production networks.
https://t.co/P21rh8eN2E

[adithya @00aleph00]

a fascinating take by William Thurston on the effect of math competitions!

[Chong Liu @ChongLiuCS]

I'm looking for two #MachineLearning #PhD students to join my lab at @UAlbany #ComputerScience starting Fall 2024!

Great team, environment, and location!

Detailed information can be found: https://t.co/TZlwI3NJEK

Please spread the word! DM me if any questions!

[Alex Kontorovich @AlexKontorovich]

"Voices from the Bombed Universities of Ukraine" by Masha Vlasenko and Efim Zelmanov, in latest @amermathsoc Notices: https://t.co/0v0vJIhUiS

[Fang Han @johnleibniz]

Look at what I found on arXiv today 🤩 ⁦@pengding00⁩ https://t.co/xGBUBDo1v2

[Saloni @salonium]

New post by me @OurWorldInData!

How do researchers study the prevalence of mental illnesses worldwide? What are the limitations with these estimates?

Thread.
https://t.co/U3TasgO6xy

[JungHo Lee @leejunghoj]

#acic2023 was so much fun! Thanks @PanosToulis for taking a picture of me with the most interesting slide of our presentation.

[Stephanie Hicks, PhD @stephaniehicks]

I am truly excited to announce a new award launched in 2023 called the JASA #Reproducibility Award, which aims to recognize outstanding papers in the Journal of the American Statistical Association (JASA) in terms of computational reproducibility!

🔗 https://t.co/7YPp6BYLlK

[Edward Kennedy @edwardhkennedy]

Very excited about this new paper!

We study if one can improve popular semiparametric / doubly robust / DML causal effect estimators -- without adding structural assumptions...

Short answer: nope!

Turns out these methods are minimax optimal here

Lots of Qs left to explore...