<s>
Guided	B-Algorithm
local	I-Algorithm
search	I-Algorithm
is	O
a	O
metaheuristic	B-Algorithm
search	O
method	O
.	O
</s>
<s>
A	O
meta-heuristic	O
method	O
is	O
a	O
method	O
that	O
sits	O
on	O
top	O
of	O
a	O
local	B-Algorithm
search	I-Algorithm
algorithm	I-Algorithm
to	O
change	O
its	O
behavior	O
.	O
</s>
<s>
Guided	B-Algorithm
local	I-Algorithm
search	I-Algorithm
builds	O
up	O
penalties	O
during	O
a	O
search	O
.	O
</s>
<s>
It	O
uses	O
penalties	O
to	O
help	O
local	B-Algorithm
search	I-Algorithm
algorithms	I-Algorithm
escape	O
from	O
local	O
minima	O
and	O
plateaus	O
.	O
</s>
<s>
When	O
the	O
given	O
local	B-Algorithm
search	I-Algorithm
algorithm	I-Algorithm
settles	O
in	O
a	O
local	O
optimum	O
,	O
GLS	O
modifies	O
the	O
objective	O
function	O
using	O
a	O
specific	O
scheme	O
(	O
explained	O
below	O
)	O
.	O
</s>
<s>
Then	O
the	O
local	B-Algorithm
search	I-Algorithm
will	O
operate	O
using	O
an	O
augmented	O
objective	O
function	O
,	O
which	O
is	O
designed	O
to	O
bring	O
the	O
search	O
out	O
of	O
the	O
local	O
optimum	O
.	O
</s>
<s>
The	O
choice	O
of	O
solution	O
features	O
depends	O
on	O
the	O
type	O
of	O
problem	O
,	O
and	O
also	O
to	O
a	O
certain	O
extent	O
on	O
the	O
local	B-Algorithm
search	I-Algorithm
algorithm	I-Algorithm
.	O
</s>
<s>
When	O
the	O
local	B-Algorithm
search	I-Algorithm
algorithm	I-Algorithm
returns	O
a	O
local	O
minimum	O
x	O
,	O
GLS	O
penalizes	O
all	O
those	O
features	O
(	O
through	O
increments	O
to	O
the	O
penalty	O
of	O
the	O
features	O
)	O
present	O
in	O
that	O
solution	O
which	O
have	O
maximum	O
utility	O
,	O
,	O
as	O
defined	O
below	O
.	O
</s>
<s>
GLS	O
uses	O
an	O
augmented	O
cost	O
function	O
(	O
defined	O
below	O
)	O
,	O
to	O
allow	O
it	O
to	O
guide	O
the	O
local	B-Algorithm
search	I-Algorithm
algorithm	I-Algorithm
out	O
of	O
the	O
local	O
minimum	O
,	O
through	O
penalising	O
features	O
present	O
in	O
that	O
local	O
minimum	O
.	O
</s>
<s>
has	O
described	O
an	O
extended	O
guided	B-Algorithm
local	I-Algorithm
search	I-Algorithm
(	O
EGLS	O
)	O
which	O
utilises	O
random	O
moves	O
and	O
an	O
aspiration	O
criterion	O
designed	O
specifically	O
for	O
penalty	O
based	O
schemes	O
.	O
</s>
<s>
The	O
resulting	O
algorithm	O
improved	O
the	O
robustness	O
of	O
GLS	O
over	O
a	O
range	O
of	O
parameter	O
settings	O
,	O
particularly	O
in	O
the	O
case	O
of	O
the	O
quadratic	B-Algorithm
assignment	I-Algorithm
problem	I-Algorithm
.	O
</s>
<s>
1992	O
)	O
and	O
based	O
partly	O
on	O
GENET	O
for	O
constraint	B-Application
satisfaction	I-Application
and	O
optimisation	O
,	O
has	O
also	O
been	O
implemented	O
in	O
the	O
.	O
</s>
<s>
extended	O
guided	B-Algorithm
local	I-Algorithm
search	I-Algorithm
to	O
multi-objective	O
optimization	O
,	O
and	O
demonstrated	O
its	O
use	O
in	O
staff	O
empowerment	O
in	O
scheduling	O
.	O
</s>
<s>
It	O
was	O
designed	O
for	O
constraint	B-Application
satisfaction	I-Application
.	O
</s>
<s>
Tabu	B-Algorithm
search	I-Algorithm
is	O
a	O
class	O
of	O
search	O
methods	O
which	O
can	O
be	O
instantiated	O
to	O
specific	O
methods	O
.	O
</s>
<s>
GLS	O
can	O
be	O
seen	O
as	O
a	O
special	O
case	O
of	O
Tabu	B-Algorithm
search	I-Algorithm
.	O
</s>
<s>
By	O
sitting	O
GLS	O
on	O
top	O
of	O
genetic	B-Algorithm
algorithm	I-Algorithm
,	O
Tung-leng	O
Lau	O
introduced	O
the	O
guided	O
genetic	O
programming	O
(	O
GGA	O
)	O
algorithm	O
.	O
</s>
