Difference from inductive
inference (2)
l
LCM
•
Functions such as
f
in
$
y.A(x,f,y)
may be
arbitrary
functions without any
finite representations
•
The object to be learned (infered) is
any attribute related to the function
f
•