Mathematics,
Год журнала:
2024,
Номер
12(10), С. 1596 - 1596
Опубликована: Май 20, 2024
This
paper
presents
a
novel
variant
of
the
teaching–learning-based
optimization
algorithm,
termed
BLTLBO,
which
draws
inspiration
from
blended
learning
model,
specifically
designed
to
tackle
high-dimensional
multimodal
complex
problems.
Firstly,
perturbation
conditions
in
“teaching”
and
“learning”
stages
original
TLBO
algorithm
are
interpreted
geometrically,
based
on
search
capability
is
enhanced
by
adjusting
range
values
random
numbers.
Second,
strategic
restructuring
has
been
ingeniously
implemented,
dividing
into
three
distinct
phases:
pre-course
self-study,
classroom
learning,
post-course
consolidation;
this
structural
reorganization
crossover
strategy
self-learning
phase
effectively
enhance
global
TLBO.
To
evaluate
its
performance,
BLTLBO
was
tested
alongside
seven
distinguished
variants
thirteen
functions
CEC2014
suite.
Furthermore,
two
excellent
algorithms
were
added
comparison
mode
five
scalable
CEC2008
The
empirical
results
illustrate
algorithm’s
superior
efficacy
handling
challenges.
Finally,
portfolio
problem
successfully
addressed
using
thereby
validating
practicality
effectiveness
proposed
method.
Mathematics,
Год журнала:
2022,
Номер
10(10), С. 1696 - 1696
Опубликована: Май 16, 2022
Remora
Optimization
Algorithm
(ROA)
is
a
recent
population-based
algorithm
that
mimics
the
intelligent
traveler
behavior
of
Remora.
However,
performance
ROA
barely
satisfactory;
it
may
be
stuck
in
local
optimal
regions
or
has
slow
convergence,
especially
high
dimensional
complicated
problems.
To
overcome
these
limitations,
this
paper
develops
an
improved
version
called
Enhanced
(EROA)
using
three
different
techniques:
adaptive
dynamic
probability,
SFO
with
Levy
flight,
and
restart
strategy.
The
EROA
tested
two
benchmarks
seven
real-world
engineering
statistical
analysis
experimental
results
show
efficiency
EROA.
Sustainability,
Год журнала:
2023,
Номер
15(14), С. 11454 - 11454
Опубликована: Июль 24, 2023
In
the
present
scenario,
air
quality
prediction
(AQP)
is
a
complex
task
due
to
high
variability,
volatility,
and
dynamic
nature
in
space
time
of
particulates
pollutants.
Recently,
several
nations
have
had
poor
emission
particulate
matter
(PM2.5)
that
affects
human
health
conditions,
especially
urban
areas.
this
research,
new
optimization-based
regression
model
was
implemented
for
effective
forecasting
pollution.
Firstly,
input
data
were
acquired
from
real-time
Beijing
PM2.5
dataset
recorded
1
January
2010
31
December
2014.
Additionally,
newer
2016
2022
four
Indian
cities:
Cochin,
Hyderabad,
Chennai,
Bangalore.
Then,
normalization
accomplished
using
Min-Max
technique,
along
with
correlation
analysis
selecting
highly
correlated
variables
(wind
direction,
temperature,
dew
point,
wind
speed,
historical
PM2.5).
Next,
important
features
selected
by
implementing
an
optimization
algorithm
named
reinforced
swarm
(RSO).
Further,
optimal
given
bi-directional
gated
recurrent
unit
(Bi-GRU)
AQP.
The
extensive
numerical
shows
proposed
obtained
mean
absolute
error
(MAE)
9.11
0.19
square
(MSE)
2.82
0.26
on
dataset.
On
both
datasets,
rate
minimal
compared
other
models.
Mathematics,
Год журнала:
2022,
Номер
10(9), С. 1567 - 1567
Опубликована: Май 6, 2022
Arithmetic
Optimization
Algorithm
(AOA)
is
a
physically
inspired
optimization
algorithm
that
mimics
arithmetic
operators
in
mathematical
calculation.
Although
the
AOA
has
an
acceptable
exploration
and
exploitation
ability,
it
also
some
shortcomings
such
as
low
population
diversity,
premature
convergence,
easy
stagnation
into
local
optimal
solutions.
The
Golden
Sine
(Gold-SA)
strong
searchability
fewer
coefficients.
To
alleviate
above
issues
improve
performance
of
AOA,
this
paper,
we
present
hybrid
with
Gold-SA
called
HAGSA
for
solving
industrial
engineering
design
problems.
We
divide
whole
two
subgroups
optimize
them
using
during
searching
process.
By
dividing
these
subgroups,
can
exchange
share
profitable
information
utilize
their
advantages
to
find
satisfactory
global
solution.
Furthermore,
used
Levy
flight
proposed
new
strategy
Brownian
mutation
enhance
algorithm.
evaluate
efficiency
work,
HAGSA,
selected
CEC
2014
competition
test
suite
benchmark
function
compared
against
other
well-known
algorithms.
Moreover,
five
problems
were
introduced
verify
ability
algorithms
solve
real-world
experimental
results
demonstrate
work
significantly
better
than
original
Gold-SA,
terms
accuracy
convergence
speed.
Informatics,
Год журнала:
2023,
Номер
10(3), С. 74 - 74
Опубликована: Сен. 18, 2023
The
utilization
of
reinforcement
learning
(RL)
within
the
field
education
holds
potential
to
bring
about
a
significant
shift
in
way
students
approach
and
engage
with
how
teachers
evaluate
student
progress.
use
RL
allows
for
personalized
adaptive
learning,
where
difficulty
level
can
be
adjusted
based
on
student’s
performance.
As
result,
this
could
result
heightened
levels
motivation
engagement
among
students.
aim
article
is
investigate
applications
techniques
determine
its
impact
enhancing
educational
outcomes.
It
compares
various
policies
induced
by
baselines
identifies
four
distinct
techniques:
Markov
decision
process,
partially
observable
deep
network,
chain,
as
well
their
application
education.
main
focus
identify
best
practices
incorporating
into
settings
achieve
effective
rewarding
To
accomplish
this,
thoroughly
examines
existing
literature
using
advance
technology.
This
work
provides
thorough
analysis
answer
questions
related
effectiveness
future
prospects.
findings
study
will
provide
researchers
benchmark
compare
usefulness
commonly
employed
algorithms
direction
research
