PLoS ONE,
Год журнала:
2023,
Номер
18(10), С. e0285631 - e0285631
Опубликована: Окт. 30, 2023
Copper
is
an
important
mineral
and
fluctuations
in
copper
prices
can
affect
the
stable
functioning
of
some
countries'
economies.
Policy
makers,
futures
traders
individual
investors
are
very
concerned
about
prices.
In
a
recent
paper,
we
use
artificial
intelligence
model
long
short-term
memory
(LSTM)
to
predict
To
improve
efficiency
model,
introduced
simulated
annealing
(SA)
algorithm
find
best
combination
hyperparameters.
The
feature
engineering
problem
AI
then
solved
by
correlation
analysis.
Three
economic
indicators,
West
Texas
Intermediate
Oil
Price,
Gold
Price
Silver
which
highly
correlated
with
prices,
were
selected
as
inputs
be
used
training
forecasting
model.
different
price
time
periods,
namely
485,
363
242
days,
chosen
for
forecasts.
forecast
errors
0.00195,
0.0019
0.00097,
respectively.
Compared
existing
literature,
prediction
results
this
paper
more
accurate
less
error.
research
provides
reliable
reference
analyzing
future
changes.
Buildings,
Год журнала:
2025,
Номер
15(4), С. 542 - 542
Опубликована: Фев. 10, 2025
Accurate
deformation
prediction
is
crucial
for
ensuring
the
safety
and
longevity
of
bridges.
However,
complex
fluctuations
pose
a
challenge
to
achieving
this
goal.
To
improve
accuracy,
bridge
method
based
on
bidirectional
gated
recurrent
unit
(BiGRU)
neural
network
error
correction
proposed.
Firstly,
BiGRU
model
employed
predict
data,
which
aims
enhance
modeling
capability
GRU
time-series
data
through
its
structure.
Then,
extract
valuable
information
concealed
in
error,
transformer
introduced
rectify
sequence.
Finally,
preliminary
results
are
integrated
yield
high-precision
results.
Two
datasets
collected
from
an
actual
health
monitoring
system
utilized
as
examples
verify
effectiveness
proposed
method.
The
show
that
outperforms
comparison
terms
robustness,
generalization
ability,
with
predicted
being
closer
Notably,
error-corrected
exhibits
significantly
improved
evaluation
metrics
compared
single
model.
research
findings
herein
offer
scientific
foundation
bridges’
early
warning
monitoring.
Additionally,
they
hold
significant
relevance
developing
models
deep
learning.
Resources Policy,
Год журнала:
2024,
Номер
92, С. 105040 - 105040
Опубликована: Апрель 30, 2024
This
study
aims
to
forecast
metal
futures
in
commodity
markets,
including
gold,
silver,
copper,
platinum,
palladium,
and
aluminium,
using
different
machine
deep
learning
models.
Prevalent
models
such
as
Stacked
Long-Short
Term
Memory,
Convolutional
LSTM,
Bidirectional
Support
Vector
Regressor,
Extreme
Gradient
Boosting,
Gated
Recurrent
Unit
are
utilized.
The
model
performance
is
assessed
by
multiple
factors
Root
Mean
Squared
Error,
Absolute
Percentage
Error.
stands
out
considering
simultaneously,
incorporating
both
Machine
Learning
Deep
models,
conducting
two
sets
of
experiments
with
a
full
sample
subsample
analysis.
In
addition,
it
uses
inputs
30-
60-days
periods
for
robustness
checks.
Error
values
suggest
that
efficient
on
prediction
the
future
prices.
However,
varies
significantly
influence
choice,
period,
performance.
Therefore,
suggests
constructing
theory
based
challenging.
