Gladly agree.
SCMs represent the world as a cascade of responses to changes in its state, where the output of each step serves as the input to the next for a given study unit. The probability distribution over the unit space links the SCM to data.
Gladly agree.
SCMs represent the world as a cascade of responses to changes in its state, where the output of each step serves as the input to the next for a given study unit. The probability distribution over the unit space links the SCM to data.
Both causation and covariation manifest themselves as associations.
Just because some features drive the black-box predictions doesnāt mean changing their values will actually change the outcome in real life. Kinda like that old joke ā nobody ever promised to feed the masses inā¦
Both causation and covariation manifest themselves as associations.
Just because some features drive the black-box predictions doesnāt mean changing their values will actually change the outcome in real life. Kinda like that old joke ā nobody ever promised to feed the masses inā¦
Thank you Elias for pushing our thinking this direction, and please pass my admiration to Drago for his productivityā¦
Iām wondering why did you use another LLM as the ground truth here? it just adds another abstraction⦠I mean, wouldnāt it be more convincing to estimateā¦
Thank you Elias for pushing our thinking this direction, and please pass my admiration to Drago for his productivityā¦
Iām wondering why did you use another LLM as the ground truth here? it just adds another abstraction⦠I mean, wouldnāt it be more convincing to estimateā¦
Iām thrilled to share that the Second Edition of The Book of Why will be released at the end of this year. It will include brief discussions of recent breakthroughs in causal inference, as well as some aspects of LLMs.
Join me on this next journey into the land of causality āā¦
Iād go even further than Schopenhauerās: āMan can do what he wills, but he cannot will what he wills.ā
Ideas are determined; ideas donāt determine.
Iād go even further than Schopenhauerās: āMan can do what he wills, but he cannot will what he wills.ā
Ideas are determined; ideas donāt determine.
āEvidence-based enterpriseā turns science into clickbait journalism; headlines get crowned as truth and then swung like a machete in every argument.
āEvidence-based enterpriseā turns science into clickbait journalism; headlines get crowned as truth and then swung like a machete in every argument.
Thatās not correct! šš» The assumption is much milder.
We just need to assume that the factors we condition on (or stratify on, in epi lingo) block all chains of dependency that could connect the exposure and the outcome.
A minimally sufficient adjustment set may not includeā¦
Thatās not correct! šš» The assumption is much milder.
We just need to assume that the factors we condition on (or stratify on, in epi lingo) block all chains of dependency that could connect the exposure and the outcome.
A minimally sufficient adjustment set may not includeā¦
Congratulations to @JacobJHutton on a successful defense!
What a feast for causal inference: causal diagrams, minimal adjustment sets, mediation analysis, direct and indirect effects, probability of benefit ā all in the context of cutting-edge clinical practice!
@yudapearlā¦
Exciting breakthrough from @eliasbareinboim -- a counterfactual-calculus (cfl-calculus), akin to do-calculus, for handling identification problems in Layer-3 of the Causal Hierarchy.
Exciting breakthrough from @eliasbareinboim -- a counterfactual-calculus (cfl-calculus), akin to do-calculus, for handling identification problems in Layer-3 of the Causal Hierarchy.
79K Followers 275 FollowingStudent of causal inference, human reasoning, and history of ideas, all viewed through the sharp lens of artificial intelligence.
6K Followers 4K FollowingFather in Training | Rheumatology & Drug Safety/Efficacy & Clinical Development | https://t.co/bZxRxgTqCb |https://t.co/WnBbZH5UcY
67 Followers 377 FollowingI am a widow with four orphaned sons who lost their father in the Gaza war. We have no source of income and we are waiting for you to share the posts that are d
1K Followers 1K FollowingHusband, dad, astute observer, occultist. Doing AI research and development for fun. I hold unpopular opinions because the truth isn't ran by consensus.
2K Followers 2K FollowingAssistant professor in computer science, University of Toronto | @UofTRobotics @VectorInst | Working on robotics, vision, and machine learning.
439 Followers 2K FollowingMarketing Professor at FGV-EAESP #analytics #marketresearch #phdchat #sciencewriting #AI Opinions my own | ā¤ļø or š ā endorsement
1K Followers 686 FollowingPhD student @Columbia University. Supervised by Jackie Gottlieb and Josh Jacobs. Trained to be physicist but ended up studying Brain.
79K Followers 275 FollowingStudent of causal inference, human reasoning, and history of ideas, all viewed through the sharp lens of artificial intelligence.
23K Followers 1K FollowingOfficial acct. @Columbia's Mailman School of Public Health is advancing evidence-based public health science and education. #HealthIsAHumanRight
951 Followers 1K FollowingEpidemiologist interested in causal inference, health econ, and technology. I am the Director of Real-World Evidence at Walmart Healthcare Research Institute
3K Followers 443 FollowingAsst. Prof. at Johns Hopkins CS.
Research on causal inference, causal discovery, generative AI, info theory, online learning.
15 Followers 8 FollowingCS PhD student at Columbia University, studying principles of causal generative models and representation learning.
Webpage: https://t.co/7wcvR5fYdo
2K Followers 2K FollowingLSE Ph.D. Prof at Center for Behavioral and Implementation Science at NUS. Lived and worked in many places. Retweet ā endorsements.
2K Followers 115 FollowingJiann-Ping Hsu/Karl E Peace Professor in Biostatistics & Statistics @UCBerkeley. Targeted (machine) Learning for #causalinference in #datascience with @tlverse.
138K Followers 21K FollowingChange facilitator, strategist, speaker. Prof. of Practice in Health & Care Improvement @WarwickBSchool. Advisor @HorizonsNHS. Senior Fellow @TheIHI. Views own.
1K Followers 1K FollowingKadoorie Associate Professor of Emergency Medicine @UniofOxford; Consultant in Emergency Medicine @OUHospitals; Fellow @GreenTempleton; Doctor @TVAirAmb.
2K Followers 2K FollowingAI research @IsomorphicLabs
Former: founder @Apple Health AI Zurich, Inaugural postdoc @VectorInst, PhD @UvA_Amsterdam. Working on applied problems in health
1K Followers 2K FollowingAssociate Professor at Department of Biostatistics @ Johns Hopkins University, interested in #wearables & #implantables data, national surveys.
1K Followers 376 FollowingCausal reasoning and decision making, prof at UCL; New book- Explaining the evidence https://t.co/jsQwkdsNBV https://t.co/UakO9pq5Ne
262 Followers 431 FollowingPT, MSc. PhD student
Clinician and researcher in hip fracture, older adults with signs of cognitive impairment, outcome measures, and outpatient interventions
22K Followers 6K FollowingTaking Hayek Seriously provides an ongoing seminar on the science & ideas of Friedrich Hayek, along with the latest research related to Hayek's works.
4K Followers 308 FollowingCell Tree Age/Rings co-inventor, PhD in geroscience, math-oriented Somatic Evolutionary, AgeCurve founder, Budapester in Cambridge, the meaning of life is life
937 Followers 634 FollowingSerial software #entrepreneur. Founder/CEO of Geminos - "AI That Knows Why" #AI #IIOT #causality. Previous IPO and sale to MSFT. Englishman in California.
No recent Favorites. New Favorites will appear here.