search

uchicago-computation-workshop / tor_wager

Repository for Tor Wager's presentation at the CSS Workshop (2/21/2019)
1 stars 0 forks source link

BayesWatch: A Matter of Opinions #4

Open policyglot opened 5 years ago

policyglot commented 5 years ago

This has been for me one of the most eye-opening applications of how Bayesian methods can complement reinforcement learning. I'll be sure to go through the source data you've linked and understand how you used it. For now, a larger philosophical question: I was going through Kaplan's book on 'Bayesian Methods in the Social Sciences' , Pg 294 (linked below), and looking into the idea of 'beliefs' (which came up frequently in your paper too.) The author says:

The subjectivist school, advocated by de Finetti, Savage, and others, allows for personal opinion to be elicited and incorporated into a Bayesian analysis. In the extreme, the subjectivist school would place no restriction on the source, reliability, or validity of the elicited opinion. The objectivist school, advocated by Jeffreys, Jaynes, Berger, Bernardo, and others, views personal opinion as in the realm of psychology, with no place in a statistical analysis.

How do you think quantitative psychology can going forward clearly define how 'beliefs' can be made more 'objective' and 'rigorous' for future use of Bayesian research methods?

https://www.guilford.com/books/Bayesian-Statistics-for-the-Social-Sciences/David-Kaplan/9781462516513

w4rner commented 5 years ago

Mr Pundit, I'm done for with your un-funny puns. If you've won the one-hundred I'll shun this free lunch and go have a ton of fun with yo Mom.