On Thu, Sep 18, 2014 at 2:10 PM, Piaget Modeler via AGI <[email protected]> 
wrote:
>
> But how do you represesent the fact that neither P nor (not P)  are in the KB?

By omitting any mention of P in the KB. But I guess what you want to
know is how you should make inferences about P. The correct answer is
"I don't know". But if you are forced to guess (such as in a
predictive model or a data compression program), then the usual way to
handle this is to fall back to a smaller or more general context for
which you have collected some statistics. In a context mixing
algorithm where you have multiple, independent predictors, you would
combine them by weighted averaging in the logistic domain (log
p/(1-p)). Omitting P from the calculation is equivalent to p(P) = 0.5,
which is 0 in the logistic domain.

-- 
-- Matt Mahoney, [email protected]


-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657
Powered by Listbox: http://www.listbox.com

Reply via email to