I've been reading articles and books on this topic and it makes me rethink why I decide to devout myself into the stats world. The debates are still going on and we can see several big names making great comments. Thanks to Prof. Gelman and technology, we don't have to travel across the country to listen to the inspirational lectures.
All starts from Andrew Gelman's blog entries and still continues.
1. Revolving disputes between J. Pearl and D. Rubin on causal inference (link1)
2. More on Pearl's and Rubin's framework for causal inference (link2) (link3)
2. More on Pearl's and Rubin's framework for causal inference (link2) (link3)
3. Causal inference and Bayesian (link4)
4. Prof. J. Pearl's response (link5)
--
Recently I've started learning some Bayesian methods, and it's always fun to learn something new. But I have doubts on several issues especially about causal inference. One of the common critiques on Bayesian methods is the concern of making up data, but it's not a real issue for me. The main concern lies in making up causality from probability.
I am still new to Bayesian methods and will keep on reading more related articles to figure it out.
No comments:
Post a Comment