International Journal of Computing,
Год журнала:
2024,
Номер
unknown, С. 287 - 293
Опубликована: Июль 1, 2024
The
International
Timetabling
Competition
(ITC)
2021
focuses
on
sports
scheduling,
a
domain
intricately
connected
to
optimizing
combinatorics
problems.
Within
the
framework
of
ITC
challenge,
crucial
task
is
precisely
allocate
matches
their
designated
time
slots.
Addressing
this
challenge
involves
utilization
Adaptive
Learning
Modified
Great
Deluge
(ALMGD)
algorithm,
which
belongs
realm
hyper-heuristics.
This
algorithm
represents
an
evolutionary
step
from
foundational
great
deluge
incorporating
acceptance
mechanism
woven
with
self-adaptive
learning.
To
assess
its
efficacy,
performance
ALMGD
scrutinized
through
comparative
analysis
hill
climbing
and
algorithms.
As
result,
proposed
can
produce
solution
that
superior
comparison
algorithm.
modified
reduce
penalty
by
36%,
while
only
29%
reaches
34%.
International Journal of Computational Intelligence Systems,
Год журнала:
2024,
Номер
17(1)
Опубликована: Ноя. 18, 2024
The
Variable
Size
and
Cost
Bin
Packing
Problem
(VSCBPP)
focuses
on
minimizing
the
overall
cost
of
containers
used
to
pack
a
specified
set
items.
This
problem
has
significant
applications
across
various
fields,
including
energy,
cargo
transport,
informatics,
among
others.
Most
research
conducted
this
concentrated
enhancing
solution
methodologies.
Recently,
some
studies
have
investigated
use
fuzzy
approaches
VSCBPP,
which
allow
for
relaxation
certain
constraints.
In
paper,
we
introduce
metaheuristic
method
solving
version
facilitating
simultaneous
two
constraints:
overloading
exclusion
specific
items
from
packing
process.
Consequently,
two-dimensional
VSCBPP
enables
us
derive
range
solutions
that
present
varying
trade-offs
between
satisfaction
levels
original
We
employ
mechanisms
multi-objective
approach
maximize
degrees
while
function.
To
demonstrate
efficacy
our
proposed
solution,
utilized
well-known
evolutionary
P-metaheuristics
(Multi-Objective
Genetic
Algorithm
NSGA-II)
S-metaheuristics
Local
Search
Ulungu
Multi-Objective
Simulated
Annealing)
specifically
tailored
VSCBPP.
Computational
experiments
were
39
instances
validate
effectiveness
approach.
Stagnation
at
local
optima
represents
a
significant
challenge
in
bio–inspired
optimization
algorithms,
often
leading
to
suboptimal
solutions.
This
paper
addresses
this
issue
by
proposing
hybrid
model
that
combines
the
Orca
Predator
Algorithm
with
Deep
Q–Learning.
is
an
technique
mimics
hunting
behavior
of
orcas.
It
solves
complex
problems
exploring
and
exploiting
search
spaces
efficiently.
Q–Learning
reinforcement
learning
deep
neural
networks.
integration
aims
turn
stagnation
problem
into
opportunity
for
more
focused
effective
exploitation,
enhancing
technique’s
performance
accuracy.
The
proposed
leverages
biomimetic
strengths
identify
promising
regions
nearby
space,
complemented
fine–tuning
capabilities
navigate
these
areas
precisely.
practical
application
approach
evaluated
using
high–dimensional
Heartbeat
Categorization
Dataset,
focusing
on
feature
selection
problem.
dataset,
comprising
electrocardiogram
signals,
provided
robust
platform
testing
our
model.
Our
experimental
results
are
encouraging,
showcasing
strategy
capability
relevant
features
without
significantly
compromising
metrics
machine
models.
analysis
was
performed
comparing
improved
method
against
its
native
version
set
state–of–the–art
algorithms.
International Journal of Computing,
Год журнала:
2024,
Номер
unknown, С. 287 - 293
Опубликована: Июль 1, 2024
The
International
Timetabling
Competition
(ITC)
2021
focuses
on
sports
scheduling,
a
domain
intricately
connected
to
optimizing
combinatorics
problems.
Within
the
framework
of
ITC
challenge,
crucial
task
is
precisely
allocate
matches
their
designated
time
slots.
Addressing
this
challenge
involves
utilization
Adaptive
Learning
Modified
Great
Deluge
(ALMGD)
algorithm,
which
belongs
realm
hyper-heuristics.
This
algorithm
represents
an
evolutionary
step
from
foundational
great
deluge
incorporating
acceptance
mechanism
woven
with
self-adaptive
learning.
To
assess
its
efficacy,
performance
ALMGD
scrutinized
through
comparative
analysis
hill
climbing
and
algorithms.
As
result,
proposed
can
produce
solution
that
superior
comparison
algorithm.
modified
reduce
penalty
by
36%,
while
only
29%
reaches
34%.
