Applied Sciences,
Год журнала:
2024,
Номер
14(16), С. 7231 - 7231
Опубликована: Авг. 16, 2024
Formulating
a
mix
design
for
3D
concrete
printing
(3DCP)
is
challenging,
as
it
involves
an
iterative
approach,
wasting
lot
of
resources,
time,
and
effort
to
optimize
the
strength
printability.
A
potential
solution
formulation
through
artificial
intelligence
(AI);
however,
being
new
emerging
field,
open-source
availability
datasets
limited.
Limited
significantly
restrict
predictive
performance
machine
learning
(ML)
models.
This
research
explores
data
augmentation
techniques
like
deep
generative
adversarial
network
(DGAN)
bootstrap
resampling
(BR)
increase
available
train
three
ML
models,
namely
support
vector
(SVM),
neural
(ANN),
extreme
gradient
boosting
regression
(XGBoost).
Their
was
evaluated
using
R2,
MSE,
RMSE,
MAE
metrics.
Models
trained
on
BR-augmented
showed
higher
accuracy
than
those
DGAN-augmented
data.
The
BR-trained
XGBoost
exhibited
highest
R2
scores
0.982,
0.970,
0.972,
0.971,
0.980
cast
compressive
strength,
printed
direction
1,
2,
3,
slump
flow
respectively.
proposed
method
predicting
(mm),
cast,
anisotropic
(MPa)
can
effectively
predict
printable
concrete,
unlocking
its
full
application
in
construction
industry.
Journal of Clinical Medicine,
Год журнала:
2025,
Номер
14(2), С. 550 - 550
Опубликована: Янв. 16, 2025
The
convergence
of
Artificial
Intelligence
(AI)
and
neuroscience
is
redefining
our
understanding
the
brain,
unlocking
new
possibilities
in
research,
diagnosis,
therapy.
This
review
explores
how
AI’s
cutting-edge
algorithms—ranging
from
deep
learning
to
neuromorphic
computing—are
revolutionizing
by
enabling
analysis
complex
neural
datasets,
neuroimaging
electrophysiology
genomic
profiling.
These
advancements
are
transforming
early
detection
neurological
disorders,
enhancing
brain–computer
interfaces,
driving
personalized
medicine,
paving
way
for
more
precise
adaptive
treatments.
Beyond
applications,
itself
has
inspired
AI
innovations,
with
architectures
brain-like
processes
shaping
advances
algorithms
explainable
models.
bidirectional
exchange
fueled
breakthroughs
such
as
dynamic
connectivity
mapping,
real-time
decoding,
closed-loop
systems
that
adaptively
respond
states.
However,
challenges
persist,
including
issues
data
integration,
ethical
considerations,
“black-box”
nature
many
systems,
underscoring
need
transparent,
equitable,
interdisciplinary
approaches.
By
synthesizing
latest
identifying
future
opportunities,
this
charts
a
path
forward
integration
neuroscience.
From
harnessing
multimodal
cognitive
augmentation,
fusion
these
fields
not
just
brain
science,
it
reimagining
human
potential.
partnership
promises
where
mysteries
unlocked,
offering
unprecedented
healthcare,
technology,
beyond.
Effectively
training
deep
learning
models
relies
heavily
on
large
datasets,
as
insufficient
instances
can
hinder
model
generalization.
A
simple
yet
effective
way
to
address
this
is
by
applying
modern
augmentation
methods,
they
synthesize
new
data
matching
the
input
distribution
while
preserving
semantic
content.
While
these
methods
produce
realistic
samples,
important
issues
persist
concerning
how
well
generalize
across
different
classification
architectures
and
their
overall
impact
in
accuracy
improvement.
Furthermore,
relationship
between
dataset
size
accuracy,
determination
of
an
optimal
level,
remains
open
question
field.
Aiming
challenges,
paper,
we
investigate
effectiveness
eight
methods—StyleGAN3,
DCGAN,
SAGAN,
RandAugment,
Random
Erasing,
AutoAugment,
TrivialAugment
AugMix—throughout
several
networks
varying
depth:
ResNet18,
ConvNeXt-Nano,
DenseNet121
InceptionResNetV2.
By
comparing
performance
diverse
datasets
from
leaf
textures,
medical
imaging
remote
sensing,
assess
which
offer
superior
generalization
capability
with
no
pre-trained
weights.
Our
findings
indicate
that
tool
for
dealing
small
achieving
gains
up
17%.
