BIO Web of Conferences,
Год журнала:
2024,
Номер
130, С. 03016 - 03016
Опубликована: Янв. 1, 2024
Ensuring
the
quality
of
drinking
water
is
a
critical
public
health
issue,
especially
in
face
increasing
industrial
pollution
and
climate
change.
This
study
explores
application
machine
learning
techniques,
specifically
XGBoost,
to
predict
potability
based
on
physical
chemical
parameters.
Initial
experiments
with
baseline
XGBoost
model
demonstrated
moderate
success
classification,
particular
difficulty
accurately
identifying
potable
water,
which
often
underrepresented
dataset.To
address
this
challenge,
SMOTE
technique
was
applied
balance
dataset,
resulting
improved
recall
for
class.
Additionally,
hyperparameter
optimization
via
RandomizedSearchCV
further
enhanced
model’s
performance,
albeit
modestly.
Reducing
complexity
by
selecting
most
significant
features
led
more
efficient
reduced
computational
overhead,
while
maintaining
comparable
accuracy.The
findings
indicate
that
an
optimized
model,
supported
data
preprocessing
feature
selection,
can
be
effectively
integrated
into
automated
monitoring
systems.
Such
system
would
enable
real-time
detection
deviations,
facilitating
prompt
corrective
actions
protect
health.
approach
not
only
demonstrates
applicability
environmental
but
also
highlights
potential
broader
resource
management
safety.
2022 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES),
Год журнала:
2023,
Номер
unknown, С. 1 - 6
Опубликована: Дек. 14, 2023
This
text
presents
a
study
of
the
causes
employee
termination
in
organizations.
Using
data
such
as
age,
gender,
marital
status,
monthly
income,
education,
and
others,
an
analysis
was
conducted
using
neural
network
algorithms
(Kohonen
maps)
Decision
Trees.
The
results
showed
that
both
methods
were
suitable
for
analysis,
but
method
lowest
error
(7.4%).
allowed
to
identify
most
significant
factors
age
employees,
status
years
employment
with
company.
will
help
optimize
process
predicting
dismissal
save
resources
on
unnecessary
analyses.
The
text
discusses
the
credit
card
debt
crisis
faced
by
issuers
in
Taiwan,
which
has
led
to
a
decrease
consumer
financial
confidence
and
poses
problems
for
both
banks
cardholders.
There
is
study
conducted
using
data
on
customers
who
took
cards
order
identify
most
important
criteria
non-payers.
aims
predict
loan
repayment
ability
neural
network
selecting
significant
factors.
literature
review
highlights
use
of
networks
determining
non-payers
potential
creating
systems
warn
owners
about
risk
behavior.
methodology
involves
processing
Kohonen
map,
type
that
uses
unsupervised
learning
visualization
clustering.
Decision
Trees
Neural
Networks
are
also
used
analysis.
Analysis
maps
did
not
provide
necessary
information
due
heterogeneous
clusters.
via
resulted
high
error
percentage
17.6
%
lacked
accuracy
predicting
average
values.
through
decision
tree
yielded
lower
12.5%
identified
status
September
August
as
An
experiment
was
dividing
factors
into
logical
groups,
resulting
difference
7%
percentages
each
group.
conclusion
suggests
method
suitable
this
model,
with
key
factor
defaulters.
BIO Web of Conferences,
Год журнала:
2024,
Номер
113, С. 05009 - 05009
Опубликована: Янв. 1, 2024
This
article
is
a
study
of
the
existing
management
structure
and
organizational
operation
department
at
trucking
company.
The
purpose
was
to
identify
uneven
distribution
workload
among
employees
develop
an
ideal
business
process
model
optimize
work
department.
Using
methods
analysis,
main
functions
their
labor
intensity
were
identified
staff
calculated.
After
implementing
proposed
changes
based
on
model,
there
significant
equalization
employee
decrease
in
total
functions.
obtained
results
indicate
increase
efficiency
optimization
use
enterprise
resources.
Journal of Infrastructure Policy and Development,
Год журнала:
2024,
Номер
8(7), С. 4074 - 4074
Опубликована: Авг. 1, 2024
This
research
delves
into
the
urgent
requirement
for
innovative
agricultural
methodologies
amid
growing
concerns
over
sustainable
development
and
food
security.
By
employing
machine
learning
strategies,
particularly
focusing
on
non-parametric
algorithms,
we
explore
assessment
of
soil
suitability
use
under
conditions
drought
stress.
Through
detailed
examination
varied
datasets,
which
include
parameters
like
toxicity,
terrain
characteristics,
quality
scores,
our
study
offers
new
insights
complexities
predicting
crops.
Our
findings
underline
effectiveness
various
models,
with
decision
tree
approach
standing
out
its
accuracy,
despite
need
comprehensive
data
gathering.
Moreover,
emphasizes
promise
merging
techniques
conventional
practices
in
science,
paving
way
novel
contributions
to
studies
practical
implementations.
The
authors
of
this
work
have
developed
the
technology
manufacturing
a
casting
made
20HMFL
steel
with
help
simulation,
and
also
possible
defects
options
for
their
elimination.
object
study
is
20XMFL.
This
used
to
manufacture
parts
steam
turbines,
cylinders,
valve
bodies,
clamps,
nozzle
boxes,
caps
coils
other
turbines.
There
not
single
branch
mechanical
engineering
instrumentation
that
does
use
cast
parts.
In
engineering,
mass
about
50%
weight
machines
mechanisms,
in
machine
tools
-
80%.
due
number
advantages
foundry
production
compared
methods
producing
blanks
or
finished
products.
moulded
both
simple
very
complex
shapes,
shaped
inner
cavities
cannot
difficult
obtain
ways.
many
cases,
simplest
cheapest
way
This
chapter
focuses
on
understanding
the
trajectories
and
prospects
for
developmentDevelopment
of
inclusiveInclusive
higher
educationEducation
in
RussianRussian
universities
based
analysisAnalysis
scientific
methodological
publications,
statistical
data
results
a
series
empiricalEmpirical
studiesStudies.
The
analysesAnalysis
detail
mission
tactical
tasks
Resource
TrainingTraining
Methodological
CentersCenter,
network
designed
to
supportSupport
educationEducation.
main
observationsObservations
obtained
during
an
studyStudies
impact
distance-inclusiveInclusive
personal
characteristics
associated
with
achievement
motivation
studentsStudents
disabilitiesDisabilities
are
presented.
In
conclusionConclusion,
discusses
barriersBarriers
challenges
E3S Web of Conferences,
Год журнала:
2024,
Номер
531, С. 05019 - 05019
Опубликована: Янв. 1, 2024
This
paper
presents
the
results
of
research
enterprise
business
processes
using
analysis
and
optimization
methods.
The
applies
matrices
pairwise
comparisons
tasks
symbols,
as
well
responsibility
matrix
to
identify
weaknesses
optimize
workloads
employees.
Measures
improve
were
developed
implemented,
which
allowed
equalize
workload
between
employees
efficiency
enterprise.
In
end,
ideal
models
created
a
set
diagrams
was
for
them,
reflecting
future
state
organization.
can
be
used
work
other
enterprises
in
food
industry
their
competitiveness
market.
BIO Web of Conferences,
Год журнала:
2024,
Номер
116, С. 05003 - 05003
Опубликована: Янв. 1, 2024
The
article
discusses
key
metrics
and
analysis
methods
for
successful
enterprise
management
to
conduct
an
evaluation
of
the
effectiveness
implementing
a
new
marketing
strategy.
Key
play
crucial
role
in
assessing
These
tools
provide
with
necessary
data
information
make
informed
decisions
adjust
strategic
plans
during
implementation.
Based
on
results
analyzing
methods,
can
form
view
how
successfully
strategy
has
been
implemented
what
adjustments
may
be
achieve
its
goals.
This
allows
adapt
changing
market
conditions,
optimize
activities
improve
competitiveness
market.
