Explainable deep learning hybrid modeling framework for total suspended particles concentrations prediction
Atmospheric Environment,
Год журнала:
2025,
Номер
unknown, С. 121079 - 121079
Опубликована: Фев. 1, 2025
Язык: Английский
Enhanced forecasting method for realized volatility of energy futures prices: A secondary decomposition-based deep learning model
Engineering Applications of Artificial Intelligence,
Год журнала:
2025,
Номер
146, С. 110321 - 110321
Опубликована: Фев. 20, 2025
Язык: Английский
Half-hourly electricity price prediction model with explainable-decomposition hybrid deep learning approach
Energy and AI,
Год журнала:
2025,
Номер
unknown, С. 100492 - 100492
Опубликована: Март 1, 2025
Язык: Английский
Multi-objective electric-carbon synergy optimisation for electric vehicle charging: Integrating uncertainty and bounded rational behaviour models
Applied Energy,
Год журнала:
2025,
Номер
389, С. 125790 - 125790
Опубликована: Март 30, 2025
Язык: Английский
Deep learning for anomaly detection in traditional Tibetan timber structures: A VAE-based model with multi-head attention and LSTM encoding and state-based thresholding
Structures,
Год журнала:
2025,
Номер
77, С. 109092 - 109092
Опубликована: Май 9, 2025
Язык: Английский
Model run monitoring and parameter modification methods
Applied Mathematics and Nonlinear Sciences,
Год журнала:
2024,
Номер
9(1)
Опубликована: Янв. 1, 2024
Abstract
The
stability
and
safety
of
industrial
process
operations
have
a
decisive
impact
on
the
high-quality
development
economy
industry.
However,
traditional
model
is
difficult
to
adapt
increasingly
complex
production
process.
In
this
paper,
based
probabilistic
linear
discriminant
analysis
model,
we
construct
fault
monitoring
for
operation,
through
kernel
density
estimation,
judge
whether
statistical
indexes
exceed
control
limit
so
as
determine
operation
system
has
fault.
Using
genetic
algorithm,
parameters
are
optimized
modified
find
optimal
value
model.
performance
its
practical
application
were
analyzed
Tennessee-Istman
process,
effect
parameter
modification
was
investigated.
experiments
indicate
that
KPLDA
model’s
improves
ability
recognize
faults
with
smaller
amplitude,
only
three
minor
errors,
provides
more
accurate
reporting
data
samples.
prediction
range
basically
overlapped
actual
measurements
until
sample
point
80,
trend
gray
score
values
above
0.95
in
points
120-200
differed
slightly
from
measurements,
better
results
overall.
Язык: Английский
China Classical Poetry Art Song Market Trend Forecast and Big Data Analysis in Music Industry
Applied Mathematics and Nonlinear Sciences,
Год журнала:
2024,
Номер
9(1)
Опубликована: Янв. 1, 2024
Abstract
In
view
of
the
fact
that
Chinese
classical
poetry
and
art
songs
are
more
widely
welcomed
in
people’s
entertainment
lives,
article
conducts
research
on
its
market
development
trend.
The
PSO-Prophet-LSTM
combined
prediction
model
is
constructed
by
combining
Prophet
LSTM
neural
network
optimizing
with
PSO
algorithm.
model’s
performance
was
tested
this
paper
it
used
to
predict
music
industry
songs.
achieved
best
results
terms
comparison
adaptation,
LOSS,
RMSE
convergence
curves
accuracy.
next
five
years,
total
output
value
expanded
from
RMB
465
billion
2024
986
2028.
capital
preservation
rate,
profit
tax
rate
value,
all
keep
growing
over
years.
size
songs,
segment
expands
overall
industry.
Язык: Английский