arXiv (Cornell University),
Год журнала:
2023,
Номер
unknown
Опубликована: Янв. 1, 2023
This
study
proposes
an
Ensemble
Differential
Evolution
with
Simula-tion-Based
Hybridization
and
Self-Adaptation
(EDESH-SA)
approach
for
inven-tory
management
(IM)
under
uncertainty.
In
this
study,
DE
multiple
runs
is
combined
a
simulation-based
hybridization
method
that
includes
self-adaptive
mechanism
dynamically
alters
mutation
crossover
rates
based
on
the
success
or
failure
of
each
iteration.
Due
to
its
adaptability,
algorithm
able
handle
complexity
uncertainty
present
in
IM.
Utilizing
Monte
Carlo
Simulation
(MCS),
continuous
review
(CR)
inventory
strategy
ex-amined
while
accounting
stochasticity
various
demand
scenarios.
enables
realistic
assessment
proposed
algo-rithm's
applicability
resolving
challenges
faced
by
IM
practical
settings.
The
empirical
findings
demonstrate
potential
im-prove
financial
performance
optimize
large
search
spaces.
makes
use
testing
Ackley
function
Sensitivity
Analysis
Perturbations
investigate
how
changes
variables
affect
objective
value.
analysis
provides
valuable
insights
into
behavior
robustness
algorithm.
Journal of King Saud University - Computer and Information Sciences,
Год журнала:
2024,
Номер
36(5), С. 102068 - 102068
Опубликована: Май 21, 2024
Long
Short-Term
Memory
(LSTM)
is
a
popular
Recurrent
Neural
Network
(RNN)
algorithm
known
for
its
ability
to
effectively
analyze
and
process
sequential
data
with
long-term
dependencies.
Despite
popularity,
the
challenge
of
initializing
optimizing
RNN-LSTM
models
persists,
often
hindering
their
performance
accuracy.
This
study
presents
systematic
literature
review
(SLR)
using
an
in-depth
four-step
approach
based
on
PRISMA
methodology,
incorporating
peer-reviewed
articles
spanning
2018-2023.
It
aims
address
how
weight
initialization
optimization
techniques
can
bolster
performance.
SLR
offers
detailed
overview
across
various
applications
domains,
stands
out
by
comprehensively
analyzing
modeling
techniques,
datasets,
evaluation
metrics,
programming
languages
associated
networks.
The
findings
this
provide
roadmap
researchers
practitioners
enhance
networks
achieve
superior
results.
Archives of Computational Methods in Engineering,
Год журнала:
2023,
Номер
31(1), С. 125 - 146
Опубликована: Июль 22, 2023
Abstract
Metaheuristic
algorithms
have
applicability
in
various
fields
where
it
is
necessary
to
solve
optimization
problems.
It
has
been
a
common
practice
this
field
for
several
years
propose
new
that
take
inspiration
from
natural
and
physical
processes.
The
exponential
increase
of
controversial
issue
researchers
criticized.
However,
their
efforts
point
out
multiple
issues
involved
these
practices
insufficient
since
the
number
existing
metaheuristics
continues
yearly.
To
know
current
state
problem,
paper
analyzes
sample
111
recent
studies
so-called
new,
hybrid,
or
improved
are
proposed.
Throughout
document,
topics
reviewed
will
be
addressed
general
perspective
specific
aspects.
Among
study’s
findings,
observed
only
43%
analyzed
papers
make
some
mention
No
Free
Lunch
(NFL)
theorem,
being
significant
result
ignored
by
most
presented.
Of
studies,
65%
present
an
version
established
algorithm,
which
reveals
trend
no
longer
based
on
analogies.
Additionally,
compilation
solutions
found
engineering
problems
commonly
used
verify
performance
state-of-the-art
demonstrate
with
low
level
innovation
can
erroneously
considered
as
frameworks
years,
known
Black
Widow
Optimization
Coral
Reef
analyzed.
study
its
components
they
do
not
any
innovation.
Instead,
just
deficient
mixtures
different
evolutionary
operators.
This
applies
extension
recently
proposed
versions.
Journal of Computational Design and Engineering,
Год журнала:
2024,
Номер
11(3), С. 12 - 42
Опубликована: Апрель 10, 2024
Abstract
In
recent
years,
scholars
have
developed
and
enhanced
optimization
algorithms
to
tackle
high-dimensional
engineering
challenges.
The
primary
challenge
of
lies
in
striking
a
balance
between
exploring
wide
search
space
focusing
on
specific
regions.
Meanwhile,
design
problems
are
intricate
come
with
various
constraints.
This
research
introduces
novel
approach
called
Hippo
Swarm
Optimization
(HSO),
inspired
by
the
behavior
hippos,
designed
address
real-world
HSO
encompasses
four
distinct
strategies
based
hippos
different
scenarios:
starvation
search,
alpha
margination,
competition.
To
assess
effectiveness
HSO,
we
conducted
experiments
using
CEC2017
test
set,
featuring
highest
dimensional
problems,
CEC2022
constrained
problems.
parallel,
employed
14
established
as
control
group.
experimental
outcomes
reveal
that
outperforms
well-known
algorithms,
achieving
first
average
ranking
out
them
CEC2022.
Across
classical
consistently
delivers
best
results.
These
results
substantiate
highly
effective
algorithm
for
both
Scientific Reports,
Год журнала:
2024,
Номер
14(1)
Опубликована: Март 28, 2024
Abstract
The
original
Harris
hawks
optimization
(HHO)
algorithm
has
the
problems
of
unstable
effect
and
easy
to
fall
into
stagnation.
However,
most
improved
HHO
algorithms
can
not
effectively
improve
ability
jump
out
local
optimum.
In
this
regard,
an
integrated
(IIHHO)
is
proposed.
Firstly,
linear
transformation
escape
energy
used
by
relatively
simple
lacks
law
prey
in
actual
nature.
Therefore,
intermittent
regulator
introduced
adjust
hawks,
which
conducive
improving
search
while
restoring
prey's
rest
mechanism;
Secondly,
uncertainty
random
vector,
a
more
regular
vector
change
mechanism
instead,
attenuation
obtained
modifying
composite
function.
Finally,
scope
Levy
flight
further
clarified,
jumping
order
modify
calculation
limitations
caused
fixed
step
size,
Cardano
formula
function
size
setting
accuracy
algorithm.
First,
performance
IIHHO
analyzed
on
Computational
Experimental
Competition
2013
(CEC
2013)
test
set
compared
with
seven
evolutionary
algorithms,
convergence
value
iterative
curve
better
than
verifying
effectiveness
proposed
Second,
another
three
state
art
(SOTA)
2022
2022)
set,
experiments
show
that
still
strong
for
optimal
value.
Third,
applied
two
different
engineering
experiments.
results
minimum
cost
prove
certain
advantages
dealing
problem
space.
All
these
demonstrate
promising
numeric
applications.
Engineering Optimization,
Год журнала:
2024,
Номер
unknown, С. 1 - 39
Опубликована: Июль 24, 2024
Metaheuristic
algorithms
play
a
crucial
role
in
engineering
optimization,
as
they
can
find
the
optimal
parameter
configuration
systems.
This
article
proposes
multi-strategy
improved
seagull
optimization
algorithm
(OPSOA)
to
solve
application
problems.
Aiming
problems
of
slow
search
speed
and
low
convergence
accuracy
standard
(SOA),
four
strategies,
including
Lévy
flight
Cauchy
mutation,
were
introduced
improve
its
performance.
Comparison
shows
that
OPSOA
incomplete
are
better
than
SOA,
indicating
each
improvement
is
effective.
By
testing
benchmark
functions
CEC
2017
2022,
it
shown
has
strong
ability
solution
superior
other
terms
speed.
The
this
practical
proves
significant
advantages
solving
complex
