From: OsherD on
From Osher Doctorow

Edwin Hewitt and Karl Stromberg, "Real and Abstract Analysis,"
Springer-Verlag: N.Y. 1965 (pages 4 and 6), and Wikipedia's "Symmetric
Difference" define the "Symmetric Difference" (Operation) DELTA by:

1) A DELTA B = (AB ' ) U (A ' B) for any sets A, B.

Now recall from the last post:

2) P(CB ' ) = P(A) - P(B)

But if A, B are bounded and intersect in a set of positive
probability, then:

3) P(A) - P(B) = P(AB ' ) + P(AB) - [P(BA ' ] - P(BA) = P(AB ' ) - P
(BA ' ) = P(AB ' ) - P(A ' B)

On the other hand, taking probabilities on both sides of (1) (if
defined) yields:

4) P(A DELTA B) = P(AB ' ) + P(A ' B)

So the set CB ' , or more precisely the probability P(CB ' ), can be
regarded as the "Real Conjugate" of P(A DELTA B) where A DELTA B is
the symmetric difference of A and B. This isn't quite what the title
of this post says, but the title was a "rough approximation".

While this does not completely specify the set C, it does suggest that
P(CB ' ) and CB ' is of more than usual interest, given some of the
non-trivial properties of A DELTA B such as:

5) A(B DELTA C) = (AB) DELTA (AC)
6) A DELTA(B DELTA C) = (A DELTA B) DELTA C

Osher Doctorow