IEEE Access,
Год журнала:
2024,
Номер
12, С. 26344 - 26363
Опубликована: Янв. 1, 2024
Groundwater
spring
plays
a
crucial
role
in
human
life,
including
water
resource
management
and
planning;
therefore,
developing
accurate
prediction
models
for
groundwater
potential
mapping
is
essential.
The
objective
of
this
research
to
introduce
confirm
new
modeling
approach
based
on
TensorFlow
Deep
Neural
Networks
(TF-DNN)
multisource
geospatial
data
spatial
potential,
with
case
study
the
tropical
province
central
highland
Vietnam.
For
task,
TF-DNN
model
structure
three
hidden
layers
32
neurons
each
was
established;
therein,
Adaptive
Moment
Estimation
(ADAM)
algorithm
used
as
an
optimizer,
whereas
Rectified
Linear
Unit
(ReLU)
activation
function,
sigmoid
transfer
function.
A
database
area,
consisting
733
locations
12
influencing
factors,
prepared
ArcGIS
Pro.
Then,
it
develop
verify
model.
Decision
Tree,
Support
Vector
Machine,
Logistic
Regression,
Random
Forest,
Classification
Regression
Trees
were
benchmark
comparison.
results
demonstrate
that
proposed
(Accuracy
=
80.5%,
F-score
0.797,
AUC
0.864)
achieves
high
global
performance,
outperforming
models.
Thus,
represents
novel
effective
tool
spatially
predicting
mapping.
map
generated
has
assist
provincial
authorities
formulating
strategies
concerning
socio-economic
development.
Cambridge Prisms Water,
Год журнала:
2024,
Номер
2
Опубликована: Янв. 1, 2024
Abstract
The
aquaculture
industry
requires
good
water
quality
for
its
successful
operation
but
produces
wastes
that
can
cause
environmental
deterioration
and
pose
high
risks
to
the
sector.
Adequate
waste
treatment
recycling
are
necessary
make
a
sustainable
profitable
contribute
circular
economy.
Polluted
sources,
excess
feeding,
overstocking,
use
of
antibiotics/chemicals
harmful
algal
blooms
major
causes
low
production
in
systems.
Discharges
untreated
would
have
serious
impacts
on
receiving
bodies,
eventually
itself.
Possible
solutions
include
technological
innovations
environmentally
friendly
systems,
efficient
processes
management
improved
legislation
governance.
Environmentally
feasible
technologies
such
as
system,
integrated
multi-trophic
aquaponics
including
features
viable
options
schemes.
Best
practices
integrating
advanced
technologies,
supported
by
automation
sensors,
modeling
artificial
intelligence-internet
things
environment,
stable
value
chain.
In
general,
low-cost
impact
reduction
through
governance
crucial
achieving
sustainability
natural
management.
