Sebastian Weichwald

What is a causal model and how is it different from a "common" statistical model? 👇 Thread on a mental picture and intuition how one may think about (a subclass of) causal models and the causal discovery problem. 1/ @bttyeo @eliasbareinboim @KordingLab @EpiEllie @causalinf

What is causal about causal models and representations?. arxiv.org/abs/2501.19335

📢 PhD opportunities at the University of Copenhagen in Statistics and Insurance/Economics Maths employment.ku.dk/phd?show=163385 (apply by April 1) and in Denmark more broadly Danish Data Science Academy Fellowship ddsa.dk/phdfellowshipc… (apply by March 5)

📢 Open Positions in Stats @ University of Copenhagen math.ku.dk/english/about/… Very liveable city, kind & helpful colleagues, excellent environment w/ vibrant community – can only recommend and happy to answer questions via email ;-) Spread the word and apply by OCT 6‼️

Adjustment Identification Distance: A gadjid for Causal Structure Learning ift.tt/jF71ZDt

📢 here's a new 𝚐𝚊𝚍𝚓𝚒𝚍 for Causal Structure Learning: 𝐆raph 𝐀𝐃𝐉ustment 𝐈dentification 𝐃istances to compare DAGs or CPDAGs (!) in terms of the identifying formulas they suggest for causal effects. 💻 Try it: pip install gadjid 📚 Read it: doi.org/10.48550/arXiv… 1/

📢 PhD positions in causality/statistics/ML in Copenhagen math.ku.dk/english/about/… → beyond the named projects, several independent stipends are available – just apply with your project ideas and supervisor suggestion 💭 aicentre.dk/phd-openings Apply by April 1st 📨

The Pioneer Centre for AI invites applications for 9 PhD Positions for different projects at the Centre across our collaboratories and partner universities. 🗓 Application deadline is 22 March and 1 April, 2024. For more information: aicentre.dk/phd-openings

Why? We need 𝒄𝒂𝒖𝒔𝒂𝒍 performance metrics to guide the development of 𝒄𝒂𝒖𝒔𝒂𝒍 structure learning algorithms. Existing graph distances are non-causal, ignore causal order, or slow 🐌 We develop new algorithms for the too fast 🚴‍♀️ and too causal Ancestor- and Oset-AID 2/

Using Neuro-Symbolic AI can boost the prediction accuracy... but are you learning the right concepts? 🤖 The answer is very tricky, but it is "no" in general 😱 Introducing our latest paper, accepted at #NeurIPS2023! With @tetraduzione, @andrea_whatever, and @looselycorrect.

🚀 New Release Announcement: Pymanopt 2.2.0 🚀 Changelog: bit.ly/pymanopt_2_2_0 You can upgrade with pip: pip install -U pymanopt

Introducing our #NeurIPS2023 paper: "Causal Component Analysis" (CauCA) – a causal generalization of Independent Component Analysis (ICA), and a special case of Causal Representation Learning (CRL). 1/10 📝: arxiv.org/abs/2305.17225 💻: github.com/akekic/causal-…

what does "alignment" mean these days in #AI and #ML? in a nutshell, that big tech has a clear agenda and we shall align with it. states and governments are doing this right now (and they influence academia in a weird feedback)

'Bike riders recognise that mandatory helmet legalisation does not tackle the risk of injury at its root [and] felt that helmets dehumanise bike riders which could lead to further violence against them.' 🚨NEW STUDY🚨: sciencedirect.com/science/articl… (cartoon by @davewalker)

On Friday, 20 Oct., I will give a talk at the University of Copenhagen. I will present work on identifiable representation learning---independent influences, multiple views & Causal Component Analysis Date/Time: 20 Oct. 2023, 11:15 Place: Aud 10, Universitetsparken 5, HCØ

How (not) to solve a bottleneck. (Original from 'Movement': blackwells.co.uk/bookshop/produ…. Enhanced by @Jtrace_a_stras)

Into identifiability and causal representation learning and close to Copenhagen? 👇 📢 @luigigres is visiting Copenhagen Causality Lab 🤩 📅 Talk: Oct 20, 11:15 math.ku.dk/english/calend… 👥 get in touch, if you want to meet with Luigi Oct 19/20 @AiCentreDK @MLSectionUCPH @CogSys

@sweichwald @marcoguazzinii @__agrei__ Try it out on your own data! You can find our @rstatstweet package with vignette here: github.com/marcoguazzini/… It will also be on @Bioconductor soon :)