3 ms·
I like Eliezer's page on this much better: http://yudkowsky.net/rational/bayes http://yudkowsky.net/rational/bayes It has been discussed on HN before: http:/
by tokenadult 17y ago
I like Eliezer's page on this much better:
http://yudkowsky.net/rational/bayes http://yudkowsky.net/rational/bayes
It has been discussed on HN before:
http://news.ycombinator.com/item?id=376631 http://news.ycombinator.com/item?id=376631
- bob_carpenter 17y agoEliezer Yadkowsky's page is a nice intro to Bayes's theorem, the understanding of which is critical for understanding why the posterior is proportional to the prior times the sampling distribution. But Bayesian stats isn't just about applying Bayes's theorem. Its key feature is using probabilities for model parameters and incorporating posterior uncertainty of their values into inference.