2011/11/24 R - elists <list...@abbacomm.net>:
> i think you are realistically confused about truly "negating something"
>
> english is not your native language is it?
No, it is not ! I am not as good in english as you but I am very good
with maths and logic!
(I want someone jumps over R-elists who tried to discredited me based
on a language barrier like Karsten Bräckelmann jumped over me before!)
Maths language is unique so if someone of you dont agree with me in
what follows I can give you lectures out of the list.
I think YOU are confused about "negating something" because negating
is not the only thing and it is NOT what I am doing !
given A -> B called direct problem or direct statment
there exists:
converse: B -> A
inverse: ~A -> ~B
contrapositive: ~B -> ~A
The direct and the inverse problem are logicaly equivalent. The same
between converse and contrapositive !
Here we are dealing with direct problem and inverse problem !
You say: lets assume all non spam
A (is spam)
B (mark as spam)
if an email is spam then mark it as spam (A->B)
I say: lets assume all spam
~A(is ham)
~B(mark as ham)
if an email is ham then mark it as ham (~A -> ~B)
both are logically the same and gives same results but with the second
aproach you need to train the system with a few mails (your contacts)
with the other you need tons of mails in Bayes DB and tons of running
code.
