Aleksander Molak {'url': 'CausalPython.io'}
@AleksanderMolak
Followers
2,642
Following
1,284
Media
542
Statuses
3,674
Author of "Causal Inference & Discovery in Python" || Host at || Causal AI for Everyone
Intl
Joined September 2015
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
Puerto Rico
• 562087 Tweets
石破さん
• 95661 Tweets
49ers
• 68691 Tweets
Cowboys
• 60096 Tweets
紙の保険証
• 53521 Tweets
絵本作家
• 36737 Tweets
せなけいこさん
• 36430 Tweets
ねないこだれだ
• 29756 Tweets
Mustafa Kemal Atatürk
• 29433 Tweets
#MondayMotivation
• 28433 Tweets
Nick Bosa
• 25737 Tweets
不在者投票525人分
• 22860 Tweets
大阪・豊中市
• 22273 Tweets
車の中の段ボール箱
• 21582 Tweets
Efendiler
• 18132 Tweets
Happy New Week
• 17198 Tweets
$ICE
• 16153 Tweets
KabinetPRABOWO GIBRANkompak
• 12947 Tweets
Langsung GASSS Kerja CEPATTT
• 12345 Tweets
#リリリエラ
• 12341 Tweets
Bato
• 10121 Tweets
Pinned Tweet
"Causal Inference & Discovery in Python" is out now!
Foreword by
@AjitJaokar
(
@UniofOxford
)
We did our best to avoid mistakes in the book but some happened
Here's an errata file with known mistakes:
If you spot more, let me know
4
23
100
Looking for ways to structure your causal journey?
I just dropped a new blog post on 6 casual books that will get you from zero to advanced!
Incl. great titles from
@yudapearl
,
@causalinf
& more
#causality
#causaltwitter
#python
#machinelearning
4
59
256
What if we could discover the true causal structure from observational data?
Too good to be true?
1/n
#causaltwitter
#causality
#machinelearning
#causaldiscovery
3
44
153
Is causality necessary for trustworthy and explainable systems?
From fair bandits to counterfactual explanations and back.
1/n
#causality
#causalinference
#fairness
#machinelearning
#explainability
#trustworthyAI
#causaltwitter
5
32
150
🙅🏽♀️ Most data scientists I talked to thought it's impossible.
The truth is that causes leave traces (sometimes).
Here's how 👇🏼👇🏼👇🏼
🧵(1/n)
#causality
#causaldiscovery
#causaltwitter
#machinelearning
#python
6
25
148
🍾 Book cover reveal and pre-order announcement!
You're one of the first people on social media to see the cover of 𝗖𝗮𝘂𝘀𝗮𝗹 𝗜𝗻𝗳𝗲𝗿𝗲𝗻𝗰𝗲 𝗮𝗻𝗱 𝗗𝗶𝘀𝗰𝗼𝘃𝗲𝗿𝘆 𝗶𝗻 𝗣𝘆𝘁𝗵𝗼𝗻!
🧵 (1/n)
#causalinference
#causality
#book
#python
#causaltwitter
5
17
137
🧘🏽 Causal Reinforcement Learning in 82 slides
(Plus awesome intro and rich references)
🧵(1/n)
#causality
#causalinference
#python
#machinelearning
#reinforcementlearning
#causaltwitter
2
23
124
👩🏽🔬👩🏼🔬 Causal Inference Symposium: From Theory to Practice
The times when causality was a purely academic sport are long gone.
🧵 (1/n)
#causality
#causalinference
#python
#machinelearning
#event
#businesstrends
#ai
#causaltwitter
1
18
110
😯 Confused about causality in Python?
Here's my take on the Python's causal ecosystem
🧵 (1/n)
#causality
#causalinference
#causaldiscovery
#causaltwitter
#python
#machinelearning
#datascience
1
25
109
Causal Foundation Models?
This brand new paper gives a perspective.
1/n
#causality
#machinelearning
#foundationmodels
3
26
103
A brand new paper by Lars Lorch, Andreas Krause and Bernhard Schölkopf proposes a new formalism to talk about causality.
[Link below]
1/n
#causaltwitter
#causality
#machinelearning
3
23
104
Now available also on ArXiv
New FREE online book on causality and machine learning.
🧵 1/n
#causaltwitter
#causality
#machinelearning
#econtwitter
#causalinference
1
14
85
😯 The greatest fear that drives people away from causality is not about complexity
It's about something much simpler
🧵 (1/n)
#causality
#causalinference
#python
#machinelearning
#datascience
#casualtwitter
3
18
80
🎉 Exciting news for all Causal Pythonistas out there!
#PyMC
just got it's new causal do-operator! (code 👇🏼)
🧵 (1/n)
#python
#causality
#causalinference
#machinelearning
#bayesian
#statistics
2
18
69
🤫 Many people see similarities between Bayesian networks and structural causal models (SCMs)...
...yet many are afraid to ask about the differences.
🧵 (1/n)
#causality
#causalinference
#bayesian
#machinelearning
#causaltwitter
3
10
72
🌠 Here are 4 Python causality libraries to learn in 2023 (a thread)
(1/6)
#causaltwitter
#causality
#causalinference
#Python
#MachineLearning
3
22
66
4
11
66
"Causal Inference and Discovery in Python" is now available 🎉
& it's Amazon's
#1
Hot New Release in the category!
Thank you for ur support! 💖
Let's make the causal revolution happen!
#causality
#python
#causaltwitter
#causalinference
#causaldiscovery
2
21
62
"The 7+1 Tools of Causal Inference" by
@yudapearl
at our Causal Parrots workshop at
@RealAAAI
#aaai2024
#causaltwitter
#causality
#llms
@matej_zecevic
@dev
1
12
61
Let's break it and see!
Violating causal discovery assumptions to test robustness
1/n
#causality
#causaltwitter
#machinelearning
#causalAI
1
10
59
Some people say that causal modeling requires too many assumptions.
And it's better to use associative modeling, because it requires less .
#causality
#machinelearning
#causaltwitter
1/n
2
9
58
🍀 Here are 4 Python libraries for estimating heterogenous treatment effects
(From traditional algorithms to flexible neural networks)
🧵 (1/n)
#causaltwitter
#python
#machinelearning
#deeplearning
2
10
54
🎨 Thinking how to visualize your causal models?
Here are 3 free tools that I use daily
🧵 (1/n)
#causality
#causalinference
#causaltwitter
#python
#machinelearning
#datascience
#datavisualization
#dataviz
1
12
50
98% of data scientists work with DAGs almost every day, but they won't tell you about it.
Do you know why?
🧵 (1/n)
#causality
#causaltwitter
#machinelearning
1
11
45
LLMs and Causality?
We teamed up to give you the most recent perspectives.
🧵 (1/n)
#causaltwitter
#causality
#llm
#machinelearning
Your truly,
@matej_zecevic
@amt_shrma
@Lianhuiq
@devendratweetin
@AleksanderMolak
@kerstingAIML
2
16
45
We were taught to ignore it
Yet, it hasn't disappeared
1/n
#causality
#causalai
#causaltwitter
#machinelearning
#churn
1
3
41
💀 Causality is dead.
Here are three common reasoning patterns that drive people to wrong conclusions about causality
...and three amazing resources to level up your understanding.
🧵(1/n)
#causality
#causalinference
#python
#causaltwitter
#machinelearning
#causaldiscovery
1
9
42
Are causal models probabilistic programs?
ChiRho has an answer.
1/n
#causaltwitter
#causality
#python
#machinelearning
1
8
38
"Nothing has changed in causality since the 80s"
Here are 5 novel ideas that marry causality with generative models, conformal prediction and topology.
#causality
#causaltwitter
#python
#neurips2023
3
4
37
😲 Can GPT-4 beat your favorite Causal Discovery algorithm?
A brand new paper by
@emrek
,
@amt_shrma
@osazuwa
and
@ChenhaoTan
presents a systematic review of GPT-class models performance on a number of causal benchmarks
🧵 (1/n)
#causaltwitter
#causality
#python
#nlp
#llm
#llms
3
9
37
❓ What are the sources of causal knowledge?
- Randomized experiments
- Personal expertise / domain knowledge
- Causal discovery / structure learning
Am I missing something important?
#causality
#causaltwitter
7
5
32
Bridging ideas in São Paulo
Our dialogue made me realize how different choices can sometimes lead to very similar outcomes
#causality
#econtwitter
#machinelearning
1/n
2
1
33
🍀 Here's a great survey paper on causal machine learning and open challenges by
@jeankaddour
and colleagues from
@ucl
and
@UniofOxford
Greatings from Berlin and hope to see you today at
@PyConDE
2023!
#causality
#causaltwitter
#machinelearning
2
8
32
When recently I told my client about this, he was shocked.
The truth is that causes leave traces (sometimes).
Here's how 👇🏼👇🏼👇🏼
1/n
#causality
#causalinference
#python
#machinelearning
#causaltwitter
#statistics
#causaldiscovery
2
0
31
I adapted the code of chapter 3 of
@CausalHuber
's book to Jupyter notebook to make it more accessible for the Python community.
I introduced some minor improvements and added some visuals.
You can check it here:
0
5
31
"We improved the F1 score by staggering 45% using CatBoost!"
"Great! Do we already know which campaign we should send?"
1/n
#causaltwitter
#machinelearning
#prediction
#causality
1
1
31
🤷🏼♀️🤷🏽♂️🤷🏽♀️ What can we do when we cannot run an experiment and cannot build a valid causal graph, but want causal conclusions?
🧵 (1/3)
#causality
#machinelearning
#statistics
#dataanalysis
#datascience
#causaltwitter
2
5
30
🟠 Here's a brand new causal blog post 👇🏼
Guido Imbens & Susan Athey called this method "arguably the most important innovation in the policy evaluation literature in the last 15 years".
#causaltwitter
#causality
#python
#machinelearning
0
9
30
🤯 Most data scientists don't know that we can carry out unbiased causal inference under hidden confounding.
(Sometimes)
🧵 (1/n)
#causality
#causalinference
#python
#machinelearning
#estimation
#causaltwitter
2
8
31
Can we meaningfully talk about causality in dynamical systems?
Some people are puzzled when it comes to dynamical systems and the idea of causation.
[PREMIERE TODAY] 👇🏼
1/n
#CausalBanditsPodcast
#causality
#causaltwitter
#dynamicalsystems
#complexity
1
4
28
⭕ 𝗖𝗮𝘂𝘀𝗮𝗹𝗣𝘆 ()
A new kid on the block. CausalPy is based on
@pymc_labs
PyMC. Developed by PyMC’s prolific
@inferencelab
with contributions from
@juanitorduz
. CausalPy implements a bunch of methods to work with quasi-experimental data.
(6/6)
1
7
28
😳 Correlation does not imply causation, but does causation imply correlation?
Here's a secret 👇🏼👇🏼👇🏼
🧵
#causality
#causalinference
#python
#machinelearning
#datascience
#statistics
4
11
29
Preparing for causal 2024?
Here's my take on the Python's causal ecosystem from Q1 2023.
Which packages (if any) should we add to the mix for 2024?
1/n
4
1
27
🦜 What's new in causality and LLMs in 60 seconds.
Can LLMs be more than Causal Parrots?
🧵 (1/n)
#causalinference
#causality
#nlp
#llm
#python
#machinelearning
#causaltwitter
2
8
26
🫣 Do we need to rethink The Ladder of Causation?
🧵 (1/n)
Imagine that we take a broad look at causality and can see all causal perspectives combined...
#causality
#causalinference
#causaldiscovery
#machinelearning
#python
#deeplearning
#causaltwitter
1
4
22
I used deep learning to predict the probability of churn. More people churned.
There is a problem in industry that almost no one talks about.
🧵 (1/n)
1
7
25
👧🏽 Jane wants to perform a casual analysis in Python
Should she use NumPy and Sci-kit Learn? Or can she do better?
🧵 (1/n)
#causality
#causalinference
#python
#machinelearning
#causaltwitter
4
5
26
I picked the model with the highest R^2 and the lowest p-value.
Lost $237k in less than a decade.
🧵 (1/n)
#causality
#finance
#investing
1
2
26
👋🏼 Here are five evaluation metrics* for your causal discovery algorithms
🧵 (1/n)
#causality
#causaltwitter
#python
#machinelearning
#datascience
#causaldiscovery
1
6
24
Can normalizing flows help with hidden confounding?
Here's NeuralCSA
1/n
#causality
#causaltwitter
#machinelearning
2
5
24
Causality and dynamical systems?
Yes. Next Monday.
I'm super excited and honored to share with you next Monday (Dec 11) we're releasing a brand new episode of the
#CausalBanditsPodcast
featuring the amazing
@NaftaliWeinberger
.
#causaltwitter
#causality
1/n
3
5
25
💥 New perspectives on LLMs and Causality
.
(Freshly baked pre-prints!)
🧵 (1/n)
#causality
#causalinference
#python
#llms
#causaltwitter
2
5
24
🙋🏽♀️ Can typing help us discover causal relationships in the data more efficiently?
🧵 (1/n)
#causalinference
#causality
#causaldiscovery
#python
#machinelearning
#causaltwitter
1
6
25
3️⃣ There are at least 3 reasons why I believe you should keep learning causality in 2023.
Here's my take 👇
🧵 (1/n)
#python
#causality
#causalinference
#machinelearning
#community
#causaltwitter
1
3
24
🥷🏼 Here's a casual concept that even seasoned scientists & senior data scientists struggle with
...and 2 fresh resources that will help you master it 👇🏼👇🏼
🧵
💡 The code snippet presents an SCM model that we use in chapter 6 of my upcoming causal book
1
3
23
The room was almost completely full until the last minute.
Our participants shared their thoughts and insights, sparking more engaging discussions.
A summary tomorrow.
At our Causal Parrots Workshop
@RealAAAI
#aaai2024
@devendratweetin
@matej_zecevic
3
2
23
💥 We struggled with the super-exponential explosion for years.
Today we can learn graphs continuously, but does it solve causal discovery?
Learn more 👇🏼👇🏼👇🏼
🧵 (1/n)
#causality
#causaltwitter
#python
#machinelearning
#datascience
#statistics
#causaldiscovery
1
5
21
On my way to Berlin to record new episodes of the Causal Bandits podcast, exploring a brand new causal book by
@CausalHuber
💖
#causaltwitter
#causality
1
0
21
Keep it clear.
We're deconfounding.
1/n
#causality
#causalinference
#confounding
#machinelearning
#causaltwitter
1
3
21
🎁 Here are 4 amazing free resources to study causality during the holiday season this year:
*
@CasualBrady
' YT course:
(1/4)
2
5
23
This is also one of the reasons why non-causal approaches to root cause analysis fail.
High variation is a (only) a (fallible) proxy of a factor's contribution importance.
First, many analyses (regression, PCA) assume high variance representation components are most important. But equally-important features may not carry equal variance; e.g. if a model computes easy (linear) and hard (nonlinear) tasks, the easy one dominates the representations! 3/
2
4
30
1
6
22
Just dropped a new blog post on causal discovery in Python.
💡 To learn more on causality join our fast growing Causal Python community here:
#causality
#causaldiscovery
#python
#machinelearning
#causaltwitter
#gcastle
0
7
22
Let's make Root Cause Analysis causal again!
Two ways to find the source of a problem
1/n
#rca
#causality
#causaltwitter
1
2
21
Have you ever wondered how to estimate causal effects from observational data in Python without losing weeks for implementation?
Check my new blog post at
@TDataScience
to get the answer.
#Python
#causality
#causalinference
#causaltwitter
Ready to jump-start your causal inference journey?
@AleksanderMolak
is here to help, with a clear, Python-based guide that explains how to estimate the causal effect of one variable on another variable from observational data.
1
10
31
0
10
19
⚡ BERT is biased...
...and so are other NLP methods when applied to causal tasks.
Can we do better? Learn more 👇🏼👇🏼👇🏼
🧵 (1/n)
#causality
#causalinference
#causaltwitter
#python
#machinelearning
#nlp
#bert
1
6
21
🧠 BrainBox and MicroDraw are real-time on-line collaborative tools for neuroscientists. Purely amazing! ❤️❤️❤️
@katjaQheuer
at
@BrainhackW
0
3
21
🧬 Generating synthetic data for causal discovery algorithms in 60 sec
(There's a hook though!)
🧵 (1/n)
#causality
#causalinference
#python
#machinelearning
#datascience
#causaldiscovery
#causaltwitter
1
5
20
I understood causality in just 5 tweets.
Or 5 misconceptions about causality and causal modeling to avoid
🧵 (1/n)
#causaltwitter
#causality
#machinelearning
1
0
20
The secret sauce for graphical (causal) models in Python.
You probably don't want to miss this package.
1/n
#causaltwitter
#causality
#graphicalmodels
1
2
20
90 pages of causal goodness.
With a free PDF.
1/n
#investing
#causality
#causaltwitter
#econtwitter
#machinelearning
1
3
21
🔍 Causal discovery
Adding expert knowledge in 𝟯 𝘀𝗶𝗺𝗽𝗹𝗲 𝘀𝘁𝗲𝗽𝘀 [Jupyter Notebook]
🧵 (1/n)
#causality
#causaldiscovery
#expertknowledge
#python
#machinelearning
#causaltwitter
1
7
21
Yesterday, I visited the Center for Mathematical Sciences at
@Cambridge_Uni
to talk with
@AliciaCurth
.
1/2
#CausalBanditsPodcast
1
1
20
@matej_zecevic
@Lianhuiq
@amt_shrma
@MSFTResearch
@devendratweetin
@kerstingAIML
@emrek
@yudapearl
@guyvdb
@AndrewLampinen
@micmylin
@CohrsKai
@AniketVashisht8
@oscar_clivio
@guyvdb
's talk about symbolic reasoning for LLMs has been eye-opening to many participants, and the presented idea of combining LLMs with hidden Markov models (HMMs) sparked many inspiring discussions.
11/n
2
2
20
4️⃣ Many data scientist are confused by this.
Here are four questions that will help you pick the right causal inference model for you project 👇🏼👇🏼👇🏼
🧵 (1/n)
#causality
#causalinference
#python
#machinelearning
#datascience
#statistics
1
5
18
Yesterday, in Edinburgh, we met with
@stephensenn
for a conversation about randomization, causal inference, and challenges and opportunities in clinical trials and drug development.
#causality
#causaltwitter
#rct
1/n
1
0
19
In their 2020 paper (),
@analisereal
and C. Hazlett from UCLA proposed a flexible method** that does not require assumptions on the functional form of the treatment assignment nor on the distribution of the unobserved confounders
🧵 (9/n)
1
3
20
Double trouble at AAAI (triple A I)
What a ride! A summary.
🧵 1/n
#causality
#causaltwitter
#aaai2024
@RealAAAI
2
6
21
Her work is often seen at NeurIPS, but do you know what keeps her going?
From causal models in healthcare to inconvenient questions about machine learning,
@AliciaCurth
(
@Cambridge_Uni
) loves to ask "why?".
1/n
#CausalBanditsPodcast
#causaltwitter
#ml
2
1
18
"There's no casual AI on major AI conferences."
"?"
As we approach the midpoint of
#NeurIPS
2023, it's a great time to emphasize the delicious selection of papers accepted at this year's event.
1/n
#causality
#causalAI
#machinelearning
#causaltwitter
2
4
20
🧑🏼🚀 I get this question very often from causal learners with background in R.
Yes, your favorite causal library is available in Python, my friends!
🧵 (1/n)
#causality
#causalinference
#causaldiscovery
#python
#machinelearning
#bayesian
1
1
19
🤷🏽♂️ Causes leave traces.
Sometimes. Part 2.
🧵 (1/n)
#causality
#causalinference
#python
#machinelearning
#datascience
#causaldiscovery
#causaltwitter
1
7
17
PDF (Cambridge University Press):
Print copy (Amazon):
5/5
0
2
19
💡 Want to play with ANM yourself? -> I implemented it for you -> check the notebook here:
🧵 (11/n)
1
2
19
🛢️ Data is the fuel. Causality is the engine.
Want your engine to work smoothly in the long run?
Let's think strategically about data storage and management.
🧵 (1/n)
#causality
#causalinference
#python
#causaltwitter
#machinelearning
#data
#datamanagement
#mlops
1
4
18
😱 Leaving computers on their own can be bad for you
👋🏼 Helping them can go a long way
Here's my take on 3 tools and strategies to make causal inference more efficient with human-in-the-loop
🧵 (1/n)
#causalinference
#python
#causaltwitter
#machinelearning
#causaldiscovery
1
2
19
@MattWalshBlog
@MattWalshBlog
controlling for "all relevant variables" is a bad startegy in most real-life analyses as it likely results in controlling for colliders, mediators or their descendants. Such controls will lead to arbitrary biases in the estimates, including sign flipping
1
0
18
💥 BAM! Ready to become a part of an event that bridges the gap between causal researchers & practitioners?
🧵 (1/n)
The best part? It's 100% online!
#python
#causality
#causalinference
#causaldiscovery
#machinelearning
#python
#conference
#community
#causaltwitter
1
2
18
Better model fit => more losses.
Is it possible in a $2.5 trillion industry based on 1000s of academic papers?
1/10 🧵
#causalinference
#causality
#investing
#finance
#machinelearning
#statistics
#causaltwitter
2
5
16
🧠 Work smarter not harder
Causality comes with a unique set of challenges, but sometimes causal problems can be solved efficiently based on what we already know.
🧵(1/n)
#causality
#causalinference
#python
#machinelearning
#datascience
#causaldiscovery
1
1
18
Who is he?
If you know, I might have something for you
The person in the photo (the one on the left) will be my guest in
#CausalBanditsPodcast
on Feb 5
I'll send one of the books in the image to one selected person, who:
1/n
#causaltwitter
#machinelearning
#causality
5
1
16
Need ideas for causal discovery from time series data?
Check this (works with hidden confounding).
1/🧵
#causaldiscovery
#causalinference
#machinelearning
#causaltwitter
1
2
16
⌛ This idea is so counterintuitive that most of us don't even... think about it
Proof?
🧵 (1/n)
#python
#causality
#causalinference
#machinelearning
#econometrics
#ai
#statistics
#causaltwitter
#epitwitter
#econtwitter
1
4
16
🏋🏽 Many people write about uplift, but most of them don't tell you how to deal with more complex scenarios
Here's what you can do 👇🏼👇🏼👇🏼
🧵 (1/n)
#causality
#causaltwitter
#python
#machinelearning
#datascience
#statistics
#uplift
#marketanalysis
#healthcare
2
6
16
♥️ Exciting talk on adding logical constraints to deep learning models by
@guyvdb
Guy Van den Broeck now @ Causal Workshop @
@NeurIPSConf
#causality
#causaltwitter
#deeplearning
1
5
17
I had a ton of meaningful discussions on causality & world models during the yesterday's poster session
Today we recorded a new episode of the
#CausalBanditsPodcast
with
@emrek
@MSFTResearch
Warming up for our wrkshp on Tue
@RealAAAI
#aaai2024
#causaltwitter
#MachineLearning
0
0
17
3️⃣ Three must-know facts about do-calculus
This causal formalism is very powerful. Use with care.
🧵 (1/n)
#causality
#causalinference
#python
#machinelearning
#mathematics
#causaltwitter
1
6
17
✌🏼2️⃣ There are two causal concepts that will bend your statistical mind if you haven't heard about them before.
(thrd🧵)
#causality
#causaltwitter
#Python
#machinelearning
💡 The code snippet below with comes from chapter 5 of my upcoming causal book:
1
4
18
Waiting for the dog-level intelligence check my friend
@AjitJaokar
's (
@UniofOxford
) course on developing AI apps:
#MachineLearning
1
1
14
A new revised version of
@smueller
, Ang Li &
@yudapearl
's paper is here.
Bounding individual effects by combining experimental & observational data, & structural information.
Paper:
#causality
#causalinference
#causaltwitter
0
3
16
