3
$\begingroup$

Am a software guy with no background in causal inference.

While I am now familiar with prediction techniques due to plethora of courses available online, I would like to seek recommendations from people here for causal inference.

As you might now, how prediction techniques tutorials are available in way which can be consumed by people from software background, I would like to learn causal inference techniques like Propensity Score Matching, Comparative Effetiveness Research, epidemiology analytics, effect estimates, Hazard Ratios, Odd ratios etc in a proper course which can take us from beginner to advanced curriculum. Basically, things that are used in observational healthcare research and public health fields. Is there any courses like what Udacity offers for Data Science (which only deals with Prediction problems). Am looking for a similar course which can take us from A-Z of causal analysis.

Can you guys direct me to some resources where I can learn such techniques? Doesn't really have to be a degree but can also be online tutorials or Youtube series

$\endgroup$
1
3
$\begingroup$

This is a bit of a random collection:

  • Richard McElreath's Statistical Rethinkining lecture series has substantial parts on causal inference.
  • There's also a good number of presentations available online by the authors of the "Causal inference book". They give in person courses, but I don't see a recording of those online.
  • This causality webinar has the endorsement of TWIML, which has some weight with me, but it's also not cheap.
  • Note exactly matching your question, but the bits I've looked at in this online book were quite well done.

Personally, I've not found Judea Pearl a very good experience, but that may just be personal preference and he would be a big name in the field.

$\endgroup$
1
  • $\begingroup$ thanks for the response. Useful. Upvoted. $\endgroup$
    – The Great
    Jun 22 at 13:36
2
$\begingroup$

Judea Pearl is definitely a scholar you should read if you're interested in modern causal inference. While he has written a fair amount on the subject, I've collated some of his more introductory material here for you.

He has some books:

which I would read the order given due to increasing technicality.

I'd also consider reading some of his papers, which are immediately accessible.

$\endgroup$
1
  • $\begingroup$ thanks, upvoted.. $\endgroup$
    – The Great
    Jul 10 at 22:19
1
$\begingroup$

I think A Crash Course in Causality: Inferring Causal Effects from Observational Data would be a good starting point.

looper resource list contains lots of nice links to introductory and research material, software and MOOCs. Disclaimer: I am the maintainer.

An other approach as a software engineer, one could work through examples from Python and CRAN packages out there. Specially CRAN packages, usually they come with nice documentation in the form of vignettes. See for example, MatchIt.

$\endgroup$
2
  • $\begingroup$ thanks, upvoted.. $\endgroup$
    – The Great
    Jul 10 at 22:19
  • $\begingroup$ I must encourage Pearl's framework, as DAGs are inherently in general more powerful in knowledge representation compare to flat data structures. $\endgroup$ Jul 11 at 1:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.