> > 4. As Kathy has emphasized, probabilistic statements are extremely
>> important for many domains, especially medicine. They can be
>> accommodated by one or more levels that are below the facts
>> in entrenchment.
>
>Hmm, that is an extremely doubtful claim. (01)
Probabilistic information ought in some cases to be extremely high up
in a "levels of entrenchment" ordering. In many cases, there just
plain isn't any such thing as necessary and sufficient conditions.
Ask any expert in machine learning or classification. The best we
can do is to list a bunch of features, and estimate a probability
distribution for the features given A, and a different probability
distribution for the features given not-A, and sometimes we also have
a base rate for A in a given context. Now if you give me an instance,
I can at the features and compute a probability for A and for not-A
given the features for the instance. (02)
In many business systems, humans have created rules for classifying
things, in order to legislate more efficient business processes.
Rules work (modulo exceptions) in business systems because we
designed them that way (and then made exceptions because it usually
turns out that exceptions are needed). But in natural systems,
necessary and sufficient conditions are more the exception than the
rule. (03)
Kathy (04)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (05)
|