Expert Systems with Applications,
Journal Year:
2023,
Volume and Issue:
219, P. 119695 - 119695
Published: Feb. 11, 2023
The
outbreak
of
the
COVID-19
pandemic
has
transpired
global
media
to
gallop
with
reports
and
news
on
novel
Coronavirus.
intensity
chatter
various
aspects
pandemic,
in
conjunction
sentiment
same,
accounts
for
uncertainty
investors
linked
financial
markets.
In
this
research,
Artificial
Intelligence
(AI)
driven
frameworks
have
been
propounded
gauge
proliferation
towards
Indian
stock
markets
through
lens
predictive
modelling.
Two
hybrid
frameworks,
UMAP-LSTM
ISOMAP-GBR,
constructed
accurately
forecast
daily
prices
10
companies
different
industry
verticals
using
several
systematic
indices
related
alongside
orthodox
technical
indicators
macroeconomic
variables.
outcome
rigorous
exercise
rationalizes
utility
monitoring
relevant
worldwide
India.
Additional
model
interpretation
Explainable
AI
(XAI)
methodologies
indicates
that
a
high
quantum
overall
hype,
coverage,
fake
news,
etc.,
leads
bearish
market
regimes.
China Finance Review International,
Journal Year:
2024,
Volume and Issue:
14(4), P. 759 - 790
Published: Jan. 4, 2024
Purpose
Owing
to
highly
volatile
and
chaotic
external
events,
predicting
future
movements
of
cryptocurrencies
is
a
challenging
task.
This
paper
advances
granular
hybrid
predictive
modeling
framework
for
the
figures
Bitcoin
(BTC),
Litecoin
(LTC),
Ethereum
(ETH),
Stellar
(XLM)
Tether
(USDT)
during
normal
pandemic
regimes.
Design/methodology/approach
Initially,
major
temporal
characteristics
price
series
are
examined.
In
second
stage,
ensemble
empirical
mode
decomposition
(EEMD)
maximal
overlap
discrete
wavelet
transformation
(MODWT)
used
decompose
original
time
into
two
distinct
sets
subseries.
third
long-
short-term
memory
network
(LSTM)
extreme
gradient
boosting
(XGB)
applied
decomposed
subseries
estimate
initial
forecasts.
Lastly,
sequential
quadratic
programming
(SQP)
fetch
forecast
by
combining
Findings
Rigorous
performance
assessment
outcome
Diebold-Mariano’s
pairwise
statistical
test
demonstrate
efficacy
suggested
framework.
The
yields
commendable
COVID-19
timeline
explicitly
as
well.
Future
trends
BTC
ETH
found
be
relatively
easier
predict,
while
USDT
difficult
predict.
Originality/value
robustness
proposed
can
leveraged
practical
trading
managing
investment
in
crypto
market.
Empirical
properties
dynamics
chosen
provide
deeper
insights.
Annals of Operations Research,
Journal Year:
2025,
Volume and Issue:
unknown
Published: Jan. 28, 2025
Abstract
The
growing
media
buzz
and
industry
focus
on
the
emergence
rapid
development
of
Metaverse
technology
have
paved
way
for
escalation
multifaceted
research.
Specific
coins
come
into
existence,
but
they
barely
seen
any
traction
among
practitioners
despite
their
tremendous
potential.
current
work
endeavors
to
deeply
analyze
temporal
characteristics
6
through
lens
predictive
analytics
explain
forecasting
process.
dearth
research
imposes
serious
challenges
in
building
model.
We
resort
a
granular
prediction
setup
incorporating
Maximal
Overlap
Discrete
Wavelet
Transformation
(MODWT)
technique
disentangle
original
series
subseries.
Facebook's
Prophet
TBATS
algorithms
are
utilized
individually
draw
predictions
components.
Aggregating
components-wise
forecasted
figures
achieve
final
forecast.
is
deployed
multivariate
setting,
applying
set
explanatory
features
covering
macroeconomic,
technical,
social
indicators.
Rigorous
performance
checks
justify
efficiency
integrated
framework.
Additionally,
interpret
black
box
typed
framework,
two
explainable
artificial
intelligence
(XAI)
frameworks,
SHAP
LIME,
used
gauge
nature
influence
predictor
variables,
which
serve
several
practical
insights.
China Finance Review International,
Journal Year:
2024,
Volume and Issue:
unknown
Published: May 30, 2024
Purpose
Stock
markets
are
essential
for
households
wealth
creation
and
firms
raising
financial
resources
capacity
expansion
growth.
Market
participants,
therefore,
need
an
understanding
of
stock
price
movements.
market
indices
individual
prices
reflect
the
macroeconomic
environment
subject
to
external
internal
shocks.
It
is
important
disentangle
impact
shocks,
uncertainty
speculative
elements
examine
them
separately
prediction.
To
aid
households,
policymakers,
paper
proposes
a
granular
decomposition-based
prediction
framework
different
time
periods
in
India,
characterized
by
states
with
varying
degrees
uncertainty.
Design/methodology/approach
Ensemble
empirical
mode
decomposition
(EEMD)
fuzzy-C-means
(FCM)
clustering
algorithms
used
decompose
into
short,
medium
long-run
components.
Multiverse
optimization
(MVO)
combine
extreme
gradient
boosting
regression
(XGBR),
Facebook
Prophet
support
vector
(SVR)
forecasting.
Application
explainable
artificial
intelligence
(XAI)
helps
identify
feature
contributions.
Findings
We
find
that
historic
volatility,
expected
uncertainty,
oscillators
variables
explain
components
their
varies
industry
state.
The
proposed
yields
efficient
predictions
even
during
COVID-19
pandemic
Russia–Ukraine
war
period.
Efficiency
measures
indicate
robustness
approach.
suggest
large-cap
stocks
relatively
more
predictable.
Research
limitations/implications
on
Indian
markets.
Future
work
will
extend
it
other
products.
Practical
implications
methodology
be
practical
use
traders,
fund
managers
advisors.
Policymakers
may
useful
assessing
shocks
reducing
volatility.
Originality/value
Development
forecasting
separating
effects
explanatory
scales
periods.
Transportation Research Part E Logistics and Transportation Review,
Journal Year:
2024,
Volume and Issue:
189, P. 103686 - 103686
Published: July 23, 2024
Prediction
of
bunker
fuel
spot
prices
at
a
port
and
understanding
the
dependence
on
key
determinants
is
an
arduous
challenging
activity.
The
present
work
strives
to
analyze
temporal
spectrum
daily
Very
Low
Sulphur
Oil
(VLSFO),
critical
fuel,
in
five
European
Ports,
Amsterdam,
Antwerp,
Gothenburg,
Hamburg,
Rotterdam.
lack
prior
research
allied
domain
has
motivated
undertake
modeling
VLSFO
through
lens
applied
predictive
analytics.
Least
Square
Boosting
(LSBoost)
Facebook
Prophet
algorithms
are
used
draw
forecasts
multivariate
framework
leveraging
constructs
related
same
different
ports,
economic
indicator,
etc.
dynamics
have
been
explicitly
examined
during
Russia-Ukraine
military
conflict.
Additionally,
Explainable
Artificial
Intelligence
(XAI)
frameworks
demystify
influence
chosen
explanatory
variables
granular
scale.
overall
findings
espouse
effectiveness
accurately
estimating
any
selected
heavily
depends
ports.
