Computers,
Год журнала:
2025,
Номер
14(5), С. 175 - 175
Опубликована: Май 4, 2025
This
research
explores
the
improvement
of
tsunami
occurrence
forecasting
with
machine
learning
predictive
models
using
earthquake-related
data
analytics.
The
primary
goal
is
to
develop
a
framework
that
integrates
wide
range
sources,
including
seismic,
geospatial,
and
ecological
data,
toward
improving
accuracy
lead
times
predictions.
study
employs
methods,
Random
Forest
Logistic
Regression,
for
binary
classification
events.
Data
collection
performed
Kaggle
dataset
spanning
1995–2023,
preprocessing
exploratory
analysis
identify
critical
patterns.
model
achieved
superior
performance
an
0.90
precision
0.88
compared
Regression
(accuracy:
0.89,
precision:
0.87).
These
results
underscore
Forest’s
effectiveness
in
handling
imbalanced
data.
Challenges
such
as
quality
interpretability
are
discussed,
recommendations
future
improvements
real-time
warning
systems.
Materials,
Год журнала:
2025,
Номер
18(9), С. 1920 - 1920
Опубликована: Апрель 24, 2025
Epoxy
resin
(EP)
is
a
candidate
material
for
offshore
oil
platform
safety
signs
due
to
its
excellent
corrosion
resistance
property.
However,
fabricating
EP
with
good
anti-corrosion
as
well
mechanical
properties
remains
significant
challenge.
Here,
we
report
new
modification
strategy
simultaneously
improve
the
and
performance
of
by
coupling
it
KH550
silanized
graphene
oxide
(KGO)
glass
fiber
(KGF).
KGO
KGF
were
grafted
onto
obtain
modified
material,
i.e.,
KGO/KGF/EP
composites
characterized
FITR,
XRD,
SEM,
TGA
confirm
successful
synthesis
composites.
It
shown
that
tensile
strength
adhesion
85.5
MPa
16.0
MPa,
which
are
10.3%
23.1%
higher
than
KGO/GF/EP.
Compared
KGF/EP,
potential
increased
9.9%
rate
decreased
98.8%.
Moreover,
fluid–structure
simulation
indicated
maximum
stress
was
within
criteria
under
extreme
wind
speeds,
demonstrating
great
sign
applications.
Computers,
Год журнала:
2025,
Номер
14(5), С. 175 - 175
Опубликована: Май 4, 2025
This
research
explores
the
improvement
of
tsunami
occurrence
forecasting
with
machine
learning
predictive
models
using
earthquake-related
data
analytics.
The
primary
goal
is
to
develop
a
framework
that
integrates
wide
range
sources,
including
seismic,
geospatial,
and
ecological
data,
toward
improving
accuracy
lead
times
predictions.
study
employs
methods,
Random
Forest
Logistic
Regression,
for
binary
classification
events.
Data
collection
performed
Kaggle
dataset
spanning
1995–2023,
preprocessing
exploratory
analysis
identify
critical
patterns.
model
achieved
superior
performance
an
0.90
precision
0.88
compared
Regression
(accuracy:
0.89,
precision:
0.87).
These
results
underscore
Forest’s
effectiveness
in
handling
imbalanced
data.
Challenges
such
as
quality
interpretability
are
discussed,
recommendations
future
improvements
real-time
warning
systems.