This
study
aims
to
identify
the
key
predictors
of
Multidimensional
Energy
Poverty
Index
(MEPI)
by
employing
advanced
Machine
Learning
(ML)
ensemble
methods.
Traditional
energy
poverty
research
often
relies
on
conventional
statistical
techniques,
which
limits
understanding
complex
socioeconomic
factors.
To
address
this
gap,
we
propose
an
approach
using
three
distinct
ML
models:
XGBoost-Random
Forest
(RF),
XGBoost-Multiple
Linear
Regression
(MLR),
and
XGBoost-Artificial
Neural
Network
(ANN).
These
models
are
applied
a
comprehensive
dataset
encompassing
various
indicators.
The
findings
demonstrate
that
XGBoost-RF
achieves
exceptional
accuracy
reliability,
with
RMSE
0.041,
R²
0.975,
PCC
0.992.
XGBoost-MLR
shows
superior
generalizability,
maintaining
consistent
0.845
across
both
testing
training
phases.
XGBoost-ANN
model
balances
complexity
predictive
capability,
achieving
0.056,
0.954
in
phase,
0.799
training.
Significantly,
identifies
'Education',
'Food
Consumption
Score
(FCS)',
'Household
Food
Insecurity
Access
Scale
(HFIA)',
'Dietary
Diversity
(DDS)'
as
critical
MEPI.
results
highlight
intricate
relationship
between
factors
related
food
security
education.
By
integrating
insights
from
these
policy
initiatives,
offers
promising
new
addressing
poverty.
It
highlights
importance
education,
security,
crafting
effective
interventions.
Heliyon,
Год журнала:
2024,
Номер
10(12), С. e33099 - e33099
Опубликована: Июнь 1, 2024
Maximizing
the
use
of
explosives
is
crucial
for
optimizing
blasting
operations,
significantly
influencing
productivity
and
cost-effectiveness
in
mining
activities.
This
work
explores
incorporation
machine
learning
methods
to
predict
powder
factor,
a
measure
assessing
effectiveness
explosive
deployment,
using
important
rock
characteristics.
The
goal
enhance
accuracy
factor
prediction
by
employing
methods,
namely
decision
tree
models
artificial
neural
networks.
analysis
finds
key
factors
that
have
substantial
impact
on
hence
enabling
more
accurate
planning
execution
operations.
uses
data
from
180
blast
rounds
carried
out
at
dolomite
mine
south-south
Nigeria.
It
incorporates
measures
such
as
root
mean
square
error
(RSME),
absolute
(MAE),
R-squared
(R2),
variance
accounted
(VAF)
determine
best
predicting
factor.
results
indicate
model
(MD4)
outperforms
alternative
approaches,
networks
Gaussian
Process
Regression
(GPR).
In
addition,
research
presents
an
efficient
network
equation
(MD2)
estimating
values
optimum
demonstrating
outstanding
fragmentation.
conclusion,
this
provides
significant
information
improving
prediction,
which
especially
advantageous
small-scale
RSC Advances,
Год журнала:
2024,
Номер
14(43), С. 31259 - 31273
Опубликована: Янв. 1, 2024
Addressing
global
freshwater
scarcity
requires
innovative
technological
solutions,
among
which
desalination
through
thin-film
composite
polyamide
membranes
stands
out.
The
performance
of
these
plays
a
vital
role
in
desalination,
necessitating
advanced
predictive
modeling
for
optimization.
This
study
harnesses
machine
learning
(ML)
algorithms,
including
support
vector
(SVM),
neural
networks
(NN),
linear
regression
(LR),
and
multivariate
(MLR),
alongside
their
ensemble
techniques
to
predict
enhance
average
water
flux
(AWF)
salt
rejection
(ASR)
essential
metrics
efficiency.
To
ensure
model
interpretability
feature
importance
analysis,
SHapley
Additive
exPlanations
(SHAP)
were
employed,
providing
both
local
insights
into
contributions.
Initially,
the
individual
models
validated,
with
NN
demonstrating
superior
AWF
ASR,
achieving
lowest
mean
absolute
error
(MAE
=
0.001)
root
squared
(RMSE
0.0111)
an
MAE
0.0107
RMSE
0.0982
ASR.
accuracy
predictions
improved
significantly
models,
as
evidenced
by
near-perfect
Nash-Sutcliffe
efficiency
(NSE)
values.
Specifically,
(NN-E)
Linear
Regression
(LR-E)
reached
0.001
0.0111,
respectively,
AWF.
For
NN-E
reduced
0.0013
0.0089,
while
LR-E
maintained
competitive
0.0133
0.0936.
SHAP
analysis
revealed
that
features
such
MDP
TMC
critical
drivers
performance,
showing
most
significant
positive
impact
on
These
findings
demonstrate
dominance
methods
over
algorithms
predicting
key
parameters.
enhanced
precision
estimating
ASR
offered
neuro-intelligent
ensembles,
combined
provided
can
lead
environmental
operational
improvements
membrane
optimizing
resource
usage
minimizing
ecological
impacts.
paves
way
integrating
intelligent
ML
ensembles
SHAP-based
practical
field
technology,
marking
step
forward
toward
sustainable
efficient
processes.
Research Square (Research Square),
Год журнала:
2025,
Номер
unknown
Опубликована: Апрель 24, 2025
Abstract
Interfacial
tension
(IFT)
between
two
immiscible
phases
is
a
key
parameter
in
various
oil
and
gas
industries,
especially
enhanced
recovery
(EOR)
Carbon
dioxide
capture
storage
(CCS).
There
are
several
laboratory
methods
for
measuring
IFT,
of
which
the
pendant
drop
method
one
most
commonly
used.
This
can
be
used
both
thermodynamic
equilibrium
dynamic
approaches.
For
more
complete
study
modeling
to
investigate
process
component
exchange
determine
mechanism
equilibrium.
this
purpose,
novel
computational
algorithm
presented
that
calculates
IFT
under
(non-thermodynamic
equilibrium)
conditions
at
different
time
intervals,
where
each
step
separately
considered
Vapor–liquid
(VLE)
calculations
were
performed
using
Peng-Robinson
equation
state
(PR-EOS),
was
calculated
Parachor
model.
The
power
proposed
model
also
matching
fit
experimental
data.
Over
time,
increases,
thereby
reducing
IFT.
decreasing
continues
until
it
reaches
constant
(thermodynamic
value.
In
step,
exchangeable
components
calculated,
their
transfer
directions
determined.
results
show
rate
differed
any
time.
However,
intermediate
intense
beginning
experiment,
but
gradually,
as
passed
exchanged
phases,
decreased.
ultimately
reduces
average
molecular
weight
viscosity
over
goals
injecting
into
reservoirs.
Therefore,
changes
composition
gas,
well
properties
oil,
reach
two-phase
paper,
decreased
by
an
approximately
31%
compared
first
contact
due
exchange.
mass
about
39%
23%,
respectively.
These
justify
use
rich
injection
because
increase
mobility
during
process.
Thus,
effectively
studies
reservoirs
accurately
identify
mechanisms
reservoir
conditions.
Scientific Reports,
Год журнала:
2025,
Номер
15(1)
Опубликована: Май 13, 2025
The
accurate
determination
of
mycotoxins
in
food
samples
is
crucial
to
guarantee
safety
and
minimize
their
toxic
effects
on
human
animal
health.
This
study
proposed
the
use
a
support
vector
regression
(SVR)
predictive
model
improved
by
two
metaheuristic
algorithms
used
for
optimization
namely,
Harris
Hawks
Optimization
(HHO)
Particle
Swarm
(PSO)
predict
chromatographic
retention
time
various
mycotoxin
groups.
dataset
was
collected
from
secondary
sources
train
validate
SVR-HHO
SVR-PSO
models.
performance
models
assessed
via
mean
square
error,
correlation
coefficient,
Nash-Sutcliffe
efficiency.
outperformed
existing
methods
4-7%
both
learning
(training
testing)
phases
respectively.
By
using
optimization,
parameter
adjustment
became
more
effective,
avoiding
trapping
local
minima
improving
generalization.
These
results
demonstrate
how
machine
metaheuristics
may
be
combined
accurately
forecast
levels,
providing
useful
tool
regulatory
compliance
monitoring.
framework
perfect
commercial
quality
assurance,
testing,
extensive
programs
because
it
provides
exceptional
accuracy
resilience
predicting
times.
In
contrast
conventional
models,
effectively
manages
intricate
nonlinear
interactions,
guaranteeing
identification
while
lowering
hazards
