2003-07-15 20:00:09
| #virus from 2003-07-15 20:00:00 (showing messages 1-30) Bookmark the permanent url.
|
20:00:09 | hermit | I don't know of one. |
20:00:18 | Ophis | ...or should we just get on with tonight's chat since its 22:00 |
20:00:20 | Lucifer | Ophis, I will lend you a book with it |
20:00:29 | rhino | bayes theorem is very useful in decision theory when you want to ensure some level of service |
20:00:31 | Lucifer | Chat begins now |
20:00:33 | hermit | It is an argument which one of my Profs (A Dutchman) used in the 70s |
20:00:52 | Lucifer | Bayes Theorem... what's the (big) deal? |
20:01:03 | Ophis | Is that our topic? |
20:01:09 | Lucifer | Many of you have noticed there is a Cult of Bayes over in #sl4 |
20:01:10 | Sat | * Sat shrugs |
20:01:12 | rhino | ophis, try doomsday argument < http://www.anthropic-principle.com/preprints/cau/paradoxes.pdf |
20:01:21 | Sat | Eli seems smitten with it |
20:01:25 | Ophis | I'm reading: http://hanson.gmu.edu/nodoom.html |
20:01:35 | Lucifer | Are they (#sl4) onto something or is this a case of mass delusion? |
20:01:36 | Ophis | "Critiquing the Doomsday Argument" |
20:02:03 | Lucifer | What are the real implications of Bayes formula? |
20:02:08 | localroger | It seems to be an attempt to quantify common sense. |
20:02:09 | rhino | nin fact that link was "for" the doomsday arguiment |
20:02:10 | Sat | I know next to nothing about the theorem. can anyone explain it? |
20:02:28 | Lucifer | localroger, interesting that you should say that |
20:02:34 | rhino | nothing is wrong with bayes formula |
20:02:46 | Lucifer | I was just reading the intro to E.T Jayne's tome |
20:02:49 | rhino | it was originally formulated for probabilistic situations |
20:02:49 | hermit | Lucifer, Bayers is very useful in three circumstances. |
20:03:02 | hermit | You have known probabilities |
20:03:18 | Lucifer | Jaynes basically starts with the same assertion, it is just common sense made into calculations |
20:03:19 | hermit | Or you have a known mechanism |
20:04:09 | hermit | Or you have a good guess, and are modifying the guess as you get data in order to obtain a posterior distribution. |
20:04:12 | rhino | going from statistical evidence to a subjective degree of belief looks like a leap |
20:04:13 | Ophis | Bayes Theorem (I'm quite the neophyte here): A formula/system that allows you to re-assess your position on a given situation based on probability calculations and re-evaluations. |