Newton Downhill Optimizer for Global Optimization
Abstract
The
study
presents
the
Newton's
Downhill
Optimizer
(NDO),
a
novel
metaheuristic
algorithm
designed
to
address
challenges
of
complex,
high-dimensional,
and
nonlinear
optimization
problems.
Mathematical-Based
Algorithms
(MBAs)
are
category
algorithms
based
on
mathematical
principles.
They
widely
applied
in
numerical
computation,
symbolic
manipulation,
geometric
processing,
problems,
probabilistic
statistics,
offering
efficient
precise
solutions
complex
Inspired
by
Method,
NDO
combines
its
precision
with
downhill
strategy
stochastic
processes,
specifically
real-world
applications
benchmark
method
inspired
enhancing
capability
exploring
solution
space
escaping
local
optima.
In
tests,
demonstrated
exceptional
performance,
surpassing
majority
competing
multiple
test
suites
CEC
2017
2022.
We
conducted
comprehensive
comparison
against
14
well-established
algorithms.
These
include
mathematical-based
approaches
such
as
AOA,
SCHO,
SCA,
SABO,
NRBO,
RUN.
also
compared
it
classical
like
CMA-ES,
ABC,
DE,
PSO.
Additionally,
we
included
advanced
recently
published
WSO,
EHO,
FDB_AGDEand
GQPSO.
results
demonstrate
that
outperforms
most
these
It
exhibits
superior
convergence
speed
remarkable
stability.In
engineering
applications,
outperformed
other
reducer
design
task
step-cone
pulley
delivered
outstanding
disk
clutch
brake
tasks.
A
significant
contribution
is
application
breast
cancer
feature
selection,
tested
two
Breast
datasets.
performance
accuracy,
sensitivity,
specificity,
Matthews
Correlation
Coefficient
(MCC),
achieving
accuracy
across
This
underscores
potential
viable
tool
for
addressing
both
medical
fields.
source
codes
will
be
shared
at
https://github.com/oykc1234/NDO.

Research Square (Research Square), Год журнала: 2025, Номер unknown
Опубликована: Апрель 1, 2025
Язык: Английский