In
this
study,
our
aim
was
to
estimate
the
adsorption
potential
of
three
families
aerogels:
nanocellulose
(NC),
chitosan
(CS),
and
graphene
(G)
oxide-based
aerogels.
The
emphasized
efficiency
seek
here
concerns
oil
organic
contaminant
removal.
order
achieve
goal,
principal
component
analysis
(PCA)
used
as
a
data
mining
tool.
PCA
showed
hidden
patterns
that
were
not
possible
by
bi-dimensional
conventional
perspective.
fact,
higher
total
variance
scored
in
study
compared
with
previous
findings
(an
increase
nearly
15%).
Different
approaches
pre-treatments
have
provided
different
for
PCA.
When
whole
dataset
taken
into
consideration,
able
reveal
discrepancy
between
nanocellulose-based
aerogel
from
one
part
chitosan-based
graphene-based
aerogels
another
part.
overcome
bias
yielded
outliers
probably
degree
representativeness,
separation
individuals
adopted.
This
approach
allowed
an
64.02%
(for
dataset)
69.42%
(outliers
excluded
79.82%
only
dataset).
reveals
effectiveness
followed
high
outliers.
Processes,
Год журнала:
2024,
Номер
12(6), С. 1145 - 1145
Опубликована: Июнь 1, 2024
In
this
study,
Principal
Component
Analysis
(PCA)
was
applied
to
discern
the
underlying
trends
for
31
distinct
MFI
(Mobil
No.
5)-zeolite
membranes
of
11
textural,
chemical,
and
operational
factors
related
manufacturing
processes.
Initially,
a
comprehensive
PCA
approach
employed
entire
dataset,
revealing
moderate
influence
first
two
principal
components
(PCs),
which
collectively
accounted
around
38%
variance.
Membrane
samples
exhibited
close
proximity,
prevented
formation
any
clusters.
To
address
limitation,
subset
acquisition
strategy
followed,
based
on
findings
dataset.
This
resulted
in
an
enhanced
overall
contribution
revelation
diverse
patterns
among
considered
(total
variance
between
55%
77%).
The
segmentation
data
unveiled
robust
correlation
silica
(SiO2)
concentration
pervaporation
conditions.
Additionally,
notable
clustering
chemical
compositions
preparation
solutions
underscored
their
significant
efficacy
zeolite
membranes.
On
other
hand,
exclusive
composition
solution
noticed.
highlighted
high
efficiency
coupling
with
experimental
results
can
provide
data-driven
enhancement
MFI-type
used
ethanol/water
separation.
Journal of Marine Science and Engineering,
Год журнала:
2024,
Номер
12(8), С. 1444 - 1444
Опубликована: Авг. 21, 2024
The
selection
of
a
navy
ship
is
essential
to
guarantee
country’s
sovereignty,
deterrence
capabilities,
and
national
security,
especially
in
the
face
possible
conflicts
diplomatic
instability.
This
paper
proposes
integration
concepts
related
multi-criteria
decision
making
(MCDM)
methodology
machine
learning,
creating
Simple
Aggregation
Preferences
Expressed
by
Ordinal
Vectors—Principal
Components
(SAPEVO-PC)
method.
proposed
method
an
evolution
SAPEVO
family,
allowing
inclusion
qualitative
preferences,
adds
from
Principal
Component
Analysis
(PCA),
aiming
simplify
decision-making
process,
maintaining
precision
reliability.
We
carried
out
case
study
analyzing
32
warships
ten
quantitative
criteria,
demonstrating
practical
application
effectiveness
generated
rankings
reflected
both
subjective
perceptions
performance
data
each
ship.
innovative
with
learning
algorithm
ensures
comprehensive
robust
analyses,
facilitating
informed
strategic
decisions.
results
showed
high
degree
consistency
reliability,
top
bottom
remaining
stable
across
different
decision-makers’
perspectives.
highlights
potential
SAPEVO-PC
improve
efficiency
complex,
environments,
contributing
field
marine
science.
Gels,
Год журнала:
2024,
Номер
10(9), С. 554 - 554
Опубликована: Авг. 27, 2024
This
study
explores
the
application
of
machine
learning
techniques,
specifically
principal
component
analysis
(PCA),
to
analyze
influence
silica
content
on
physical
and
chemical
properties
aerogels.
Silica
aerogels
are
renowned
for
their
exceptional
properties,
including
high
porosity,
large
surface
area,
low
thermal
conductivity,
but
mechanical
brittleness
poses
significant
challenges.
The
initially
utilized
cross-correlation
examine
relationships
between
key
such
as
Brunauer-Emmett-Teller
(BET)
pore
volume,
density,
conductivity.
However,
weak
correlations
prompted
PCA
uncover
deeper
insights
into
data.
results
demonstrated
that
has
a
impact
aerogel
with
first
(PC1)
showing
strong
positive
correlation
(R
Journal of Engineered Fibers and Fabrics,
Год журнала:
2024,
Номер
19
Опубликована: Янв. 1, 2024
In
the
textile
industry,
distinguishing
between
wool
and
cashmere
can
be
a
challenging
task.
Extensive
research
based
on
microscopic
images
of
two
has
achieved
very
good
results.
However,
slide
preparation
process
required
for
this
approach
is
time-consuming
labor-intensive,
limiting
its
practical
application.
To
address
challenge,
paper
proposes
new
method
that
integrates
artificial
neural
networks
hyperspectral
imaging
technology.
The
novelty
lies
in
fact
it
does
not
require
sample
preparation,
more
simple,
fast,
nondestructive.
Firstly,
total
225
samples
160
were
selected
from
acquired
images.
spectral
curves
(range
900–2500
nm)
these
extracted
using
Region
Interest
(ROI)
tool
ENVI
software,
their
characteristics
analyzed.
Subsequently,
due
to
similarities
strong
correlation
curves,
Principal
Component
Analysis
(PCA)
was
employed
reduce
dimensionality
data.
A
single-layer
network
multi-layer
developed
LR
(Logistic
Regression)
MLP
(Multilayer
Perceptron)
models,
respectively,
with
training-to-validation
set
ratio
7:3.
model
trained
an
accuracy
90.3%
training
81.0%
validation
set,
suggesting
underfitting.
performed
best
five
principal
components,
attaining
94.1%
92.2%.
Precision,
recall,
F1-score
used
evaluate
comparison
classification
performance
models
revealed
significantly
outperformed
model.
Therefore,
application
technology
enables
rapid
non-destructive
identification
cashmere.
