International Journal of Architectural Engineering Technology,
Год журнала:
2024,
Номер
11, С. 124 - 139
Опубликована: Дек. 28, 2024
Accurately
predicting
equivalent
primary
energy
use
(EPEU)
in
buildings
is
crucial
for
advancing
energy-efficient
design,
optimizing
operational
strategies,
and
achieving
sustainability
goals
the
built
environment.
This
study
aims
to
develop
reliable
prediction
models
EPEU
by
leveraging
a
comprehensive
high-quality
dataset
from
Portland,
USA.
To
achieve
this,
systematic
machine
learning
framework
adopted,
encompassing
feature
selection,
data
preprocessing,
model
training,
performance
evaluation.
Several
state-of-the-art
algorithms
are
applied,
including
Random
Forest
(RF),
Gradient
Boosting
Decision
Tree
(GBDT),
Back-Propagation
Neural
Networks
(BP).
These
trained
using
key
features
such
as
building
type,
gross
floor
area,
construction
year,
various
characteristics
that
known
significantly
influence
consumption
patterns.
The
carefully
cleaned
normalized
ensure
generalizability
minimize
bias.
Model
assessed
standard
statistical
metrics,
coefficient
of
determination
(R²),
Mean
Absolute
Error
(MAE),
Root
Squared
(RMSE).
Among
tested
models,
ensemble
methods—particularly
RF
GBDT—consistently
outperform
others
terms
accuracy,
robustness,
stability
across
different
types.
results
this
not
only
highlight
potential
tasks
but
also
provide
actionable
insights
architects,
engineers,
facility
managers,
policymakers.
By
identifying
most
influential
variables
employing
effective
predictive
research
supports
data-driven
decision-making
processes
aimed
at
improving
performance.
Scientific Reports,
Год журнала:
2025,
Номер
15(1)
Опубликована: Апрель 30, 2025
Affordable
and
efficient
agricultural
methods
enhance
crop
yield
water
management
by
optimizing
resources.
Precise
irrigation
relies
on
accurate
estimation
of
reference
evapotranspiration
(ETo).
Numerous
analytical
empirical
exist
to
compute
ETo
but
these
are
costlier,
requires
time
perform
poorly
under
limited
availability
meteorological
data.
This
study
first
evaluated
the
performances
three
deep
learning
sequential
models-Long
short-term
memory
(LSTM),
Neural
Basis
Expansion
Analysis
for
Time
Series
(N-BEATS)
and,
Temporal
Convolutional
Network
model
(TCN),
predicting
daily
possessing
temporal
characteristics.
In
this
TCN
is
considered
as
baseline
be
compared
with
other
models.
results,
performed
better,
so
it
further
utilized
evaluate
two
strategies
prediction
that
makes
second
objective
paper.
approach,
historic
data
used
predict
future
using
which
standard
method.
And,
in
recursive
predicted
climatological
computed.
required
better
planning
data-scarce
situations.
The
results
demonstrate
provided
satisfactory
performance
Nash-Sutcliffe
Efficiency
(NSE)
=
0.99,
Theil
U2
0.005,
RMSE
0.092
MAE
0.048.
Also,
strategy,
values
computed
found
more
than
approach.
Thus,
comparative
among
architecture
revealed
outperformed
LSTM
N-BEATS
models
an
method
time-series
could
also
assist
precise
resources
scarcity.
Abstract
Crop
evapotranspiration
(
ET
c
)
is
a
critical
factor
for
understanding
water
demand
in
agricultural
systems,
influencing
irrigation
scheduling
and
resource
management.
Identifying
the
meteorological
factors
crucial
predicting
variations
needs
optimizing
plans.
Traditional
correlation
analysis
methods,
such
as
Pearson
correlation,
often
fail
to
capture
time‐frequency
,
which
limits
their
ability
effectively
identify
primary
factors.
This
study
integrates
Penman–Monteith
model,
analysis,
wavelet
vector
projection
length
calculation
method
propose
comprehensive
approach
identifying
secondary
influences
on
from
perspective.
Using
rice
Oryza
sativa
Gaoyou
Irrigation
District
of
Jiangsu
Province,
China,
case
study,
research
examines
seven
factors—including
air
temperature,
relative
humidity,
rainfall,
sunshine
duration—along
with
four
circulation
indices,
East
Asian
Summer
Monsoon
index
ENSO
index,
1980
2021.
The
results
indicate
that
duration
humidity
are
significant
affecting
high‐frequency
low‐frequency
signal
components
local
respectively.
Additionally,
other
factors,
minimum
show
strong
correlations
signals
within
specific
frequency
bands,
positioning
them
presents
versatile
framework
can
be
extended
areas
hydrometeorology
beyond.
Molecular & cellular biomechanics,
Год журнала:
2024,
Номер
21(3), С. 739 - 739
Опубликована: Ноя. 25, 2024
This
study
focuses
on
cross-lingual
short
text
classification
tasks
and
aims
to
combine
the
advantages
of
BERT
Multi-layer
Collaborative
Convolutional
Neural
Network
(MCNN)
build
an
efficient
model.
model
provides
rich
semantic
information
for
with
its
powerful
language
understanding
bidirectional
context
modeling
ability,
while
MCNN
effectively
extracts
local
global
features
in
through
multi-layer
convolution
structure
collaborative
working
mechanism.
In
this
study,
output
is
used
as
input
MCNN,
further
mine
deep
text,
so
realize
high-precision
text.
The
experimental
results
show
that
has
achieved
significant
performance
improvement
dataset,
which
a
new
effective
solution
tasks.
