<s>
Fitness	B-Algorithm
approximation	I-Algorithm
aims	O
to	O
approximate	O
the	O
objective	O
or	O
fitness	O
functions	O
in	O
evolutionary	O
optimization	O
by	O
building	O
up	O
machine	O
learning	O
models	O
based	O
on	O
data	O
collected	O
from	O
numerical	O
simulations	O
or	O
physical	O
experiments	O
.	O
</s>
<s>
The	O
machine	O
learning	O
models	O
for	O
fitness	B-Algorithm
approximation	I-Algorithm
are	O
also	O
known	O
as	O
meta-models	O
or	O
surrogates	O
,	O
and	O
evolutionary	O
optimization	O
based	O
on	O
approximated	O
fitness	O
evaluations	O
are	O
also	O
known	O
as	O
surrogate-assisted	O
evolutionary	O
approximation	O
.	O
</s>
<s>
Fitness	B-Algorithm
approximation	I-Algorithm
in	O
evolutionary	O
optimization	O
can	O
be	O
seen	O
as	O
a	O
sub-area	O
of	O
data-driven	O
evolutionary	O
optimization	O
.	O
</s>
<s>
Adaptive	O
fuzzy	O
fitness	O
granulation	O
(	O
AFFG	O
)	O
is	O
a	O
proposed	O
solution	O
to	O
constructing	O
an	O
approximate	O
model	O
of	O
the	O
fitness	O
function	O
in	O
place	O
of	O
traditional	O
computationally	O
expensive	O
large-scale	O
problem	O
analysis	O
like	O
(	O
L-SPA	O
)	O
in	O
the	O
Finite	B-Application
element	I-Application
method	I-Application
or	O
iterative	O
fitting	O
of	O
a	O
Bayesian	O
network	O
structure	O
.	O
</s>
<s>
This	O
granulation-based	O
fitness	B-Algorithm
approximation	I-Algorithm
scheme	O
is	O
applied	O
to	O
solve	O
various	O
engineering	O
optimization	O
problems	O
including	O
detecting	O
hidden	O
information	O
from	O
a	O
watermarked	O
signal	O
in	O
addition	O
to	O
several	O
structural	O
optimization	O
problems	O
.	O
</s>
