Abram,
> > Anyway, perhaps I can try to shed some light on the broader exchange? > My route has been to understand "A is B" as not P(A|B), but instead > P("A is X" | "B is X") plus the extensional equivalent... under this > light, the negative evidence presented by two statements "B is C" and > "A is not C" reduces the frequency of "A is B", but does not obviously > have any bearing on "B is A". (Perhaps it does have some indirect > bearing, for example through some rule of inversion... but of course > the system is not yet even well-defined, so I'll not speculate.) The way we deal with intension in PLN is a little different.. We define a fuzzy set A_PAT associated with a term A, and then the degree of membership of W in A_PAT is of the form x*y where x is a term measuring the simplicity of W relative to A and y = [P(W | A) - P(W)]^+ (where []^+ denotes the positive part) We can then measure P( A_PAT | B_PAT) and so forth... -- Ben ------------------------------------------- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244&id_secret=114414975-3c8e69 Powered by Listbox: http://www.listbox.com