Journal of Physics Conference Series,
Год журнала:
2024,
Номер
2901(1), С. 012026 - 012026
Опубликована: Ноя. 1, 2024
Abstract
In
the
realm
of
oil
exploration,
there
is
an
increasing
demand
for
precise
lithological
analysis,
particularly
in
rapid
and
accurate
identification
fine
rock
cutting
images.
Therefore,
a
novel
image
recognition
method
based
on
fusion
color
texture
features
proposed.
This
utilizes
histogram
grayscale
co-occurrence
matrix
techniques
to
extract
from
target
images,
respectively.
Compared
with
traditional
single-feature
methods,
this
integrated
feature
can
greatly
improve
accuracy
fragment
ensure
more
classification
by
designing
structure
set.
The
model
established
using
support
vector
machine
(SVM)
classifier
realize
automatic
cuttings
not
only
reduces
time
labor
intensity
manual
operation,
but
also
improves
efficiency
speed
which
meets
needs
modern
efficient
drilling
operations.
More
detailed
stratigraphic
data
be
provided
high
precision
chip
lithology
analysis.
These
have
important
reference
value
geologists
analyze
distribution,
determine
location
distribution
underground
gas
layers,
optimize
decisions
operation
plans.
Experimental
results
show
that
achieves
overall
than
90%
task
detecting
126
images
conglomerate,
153
mudstone,
150
sandstone,
up
94%
mudstone
sandstone.
It
proved
proposed
paper
better
classify
conglomerate
chips
accuracy.
Aerospace,
Год журнала:
2025,
Номер
12(2), С. 106 - 106
Опубликована: Янв. 31, 2025
The
use
of
machine
learning
techniques
to
identify
contributing
factors
in
air
incidents
has
grown
significantly,
helping
and
prevent
accidents
improve
safety.
In
this
paper,
classifier
models
such
as
LS,
KNN,
Random
Forest,
Extra
Trees,
XGBoost,
which
have
proven
effective
classification
tasks,
are
used
analyze
incident
reports
parsed
with
natural
language
processing
(NLP)
techniques,
uncover
hidden
patterns
future
incidents.
Metrics
precision,
recall,
F1-score
accuracy
assess
the
degree
correctness
predictive
models.
adjustment
hyperparameters
is
obtained
Grid
Search
Bayesian
Optimization.
KNN
had
best
rating,
followed
by
Forest
Trees.
results
indicate
that
tools
classify
helps
their
root
cause,
improving
situational
decision-making.
Frontiers in Neurorobotics,
Год журнала:
2025,
Номер
19
Опубликована: Янв. 23, 2025
Traffic
forecasting
is
crucial
for
a
variety
of
applications,
including
route
optimization,
signal
management,
and
travel
time
estimation.
However,
many
existing
prediction
models
struggle
to
accurately
capture
the
spatiotemporal
patterns
in
traffic
data
due
its
inherent
nonlinearity,
high
dimensionality,
complex
dependencies.
To
address
these
challenges,
short-term
model,
Trafficformer,
proposed
based
on
Transformer
framework.
The
model
first
uses
multilayer
perceptron
extract
features
from
historical
data,
then
enhances
spatial
interactions
through
Transformer-based
encoding.
By
incorporating
road
network
topology,
mask
filters
out
noise
irrelevant
interactions,
improving
accuracy.
Finally,
speed
predicted
using
another
perceptron.
In
experiments,
Trafficformer
evaluated
Seattle
Loop
Detector
dataset.
It
compared
with
six
baseline
methods,
Mean
Absolute
Error,
Percentage
Root
Square
Error
used
as
metrics.
results
show
that
not
only
has
higher
accuracy,
but
also
can
effectively
identify
key
sections,
great
potential
intelligent
control
optimization
refined
resource
allocation.
Maritime Business Review,
Год журнала:
2025,
Номер
unknown
Опубликована: Фев. 24, 2025
Purpose
A
combined
approach
of
additive
Holt–Winters,
support
vector
regression,
simple
moving
average
and
generalized
simulated
annealing
with
error
correction
optimal
parameter
selection
techniques
emphasizing
smoothing
period
in
residual
adjustment
is
developed
proposed
to
predict
datasets
container
throughput
at
major
ports.
Design/methodology/approach
The
Holt–Winters
model
describes
level,
trend
seasonal
patterns
provide
values
residuals.
In
addition,
the
fitted
predicts
a
future
value.
Afterwards,
series
improved
by
using
more
obvious
steady
Subsequently,
regression
formulates
nonlinear
complex
function
residuals
based
on
parameters
describe
remaining
pattern
searches
for
model.
Finally,
value
are
aggregated
be
Findings
applied
forecast
two
ports
Thailand.
empirical
results
revealed
that
outperforms
all
other
models
three
accuracy
measures
test
datasets.
still
superior
metrics
overall
additional
unseen
as
well.
Consequently,
can
useful
tool
supporting
decision-making
port
management
Originality/value
emphasizes
