
Adjunct Professor @DataSciColumbia; former Head of Research Data Science at @BarclaysIB ; formerly data @Buzzfeed;
New York, NY
Joined March 2009
- Tweets1.9K
- Following1.7K
- Followers1.8K
- Likes4.6K
Pinned Tweet
I just published “Causal Data Science” medium.com/p/causal-data-sci…
Statistical forecasting models always made me uneasy. I took some time to articulate why! There's a precise causal explanation for why they can break down.
medium.com/@akelleh/non-caus…
2
174
How other folks like to keep their technical skills sharp? I feel like it's like squeezing a balloon -- working in one area lets skill in another atrophy.
1
1
167
Inspired by a conversation here with @arpitrage and @alz_zyd_ a few weeks ago, I wanted to show where Pearl's approach to modeling systems has not yet (but should) make a huge impact in DSGE and CGE modeling. open.substack.com/pub/akelle…
Inspired by the discussion on EconTwitter about what causal inference gives us beyond prediction, I wrote a blog post with my two cents:
akelleh.substack.com/p/a-new…
I'm planning to follow with another that gets into some detail around my comments with GTAP.
Adam Kelleher retweeted
i'm pretty cynical about ai but honestly every day an llm i'm working with does something that feels really magical
Adam Kelleher retweeted
So I knew that causal inference was a big part of data science but it wasn't until I started working (again) as a data scientist that I realized
Most everything we do is causal and the question is not if we should be doing CI, but rather how much effort we should spend
Adam Kelleher retweeted
#RecSys2023 Full paper is due today, April 21. Call for short papers is still alive with abstract due on May 2 and full submission on May 9
Adam Kelleher retweeted
In 2018, "Parachute use to prevent death and major trauma when jumping from aircraft: randomized controlled trial" was published in the British Medical Journal.
It sought to determine if wearing a parachute when jumping from a plane had any impact on survivability.
It did not.
Adam Kelleher retweeted
Causal inference people, what are you using in practice for dealing with (potential) positivity/overlap violations? Do you try to detect positivity violations?
Adam Kelleher retweeted
Hello, I've been using Matplotlib for 7+ years and I still google how to do everything except plt.plot()
Adam Kelleher retweeted
People have spent $x00,000 on education and decades of their life to be in the US with established families. They’re 90%ile+ taxpayers.
Yet overnight, at the snap of a finger, 10,000+ families can be displaced.
Full article:
uscis.gov/newsroom/alerts/op…
12/12
Adam Kelleher retweeted
DoWhy 0.9 includes some exciting new extensions and features, including better sensitivity analyses, new identification algorithms, and more. I'm particularly excited to see so many new contributors joining in on this release!
#causality #causalinf #causaltwitter
Decided to join econtwitter on Mastodon! Find me at @adamkelleher@econtwitter.net
1
327
Not to alarm anyone, but ChatGPT knows d3!
You’re unable to view this Post because this account owner limits who can view their Posts. Learn more
Julia got me this great bird feeder that takes pictures and identifies birds! I’ve been loving it. This little one says hello! Their first tweet 😂 (it’s a tufted titmouse)















