Fermentation,
Journal Year:
2023,
Volume and Issue:
9(2), P. 120 - 120
Published: Jan. 26, 2023
In
this
work,
sugarcane
vinasse
combined
with
organic
waste
(food
and
wasted
tea)
was
demonstrated
to
be
an
excellent
source
of
biomethane
synthesis
from
carbon-rich
biowaste.
The
discarded
tea
trash
might
successfully
used
generate
bioenergy.
uncertainties
costs
associated
experimental
testing
were
recommended
decreased
by
the
effective
use
contemporary
machine
learning
methods
such
as
Gaussian
process
regression.
training
hyperparameters
are
crucial
in
construction
a
robust
ML-based
model.
To
make
autoregressive,
fine-tuned
employing
Bayesian
approach.
value
R2
found
greater
during
model
test
phase
0.72%,
assisting
avoidance
overtraining.
mean
squared
error
36.243
21.145
phase.
absolute
percentage
under
0.1%,
which
0.085%
throughout
model’s
research
that
combination
trash,
food
may
viable
for
generation.
methodology
approach
tuning
regression
is
efficient
method
prediction
despite
low
correlation
across
data
columns.
It
possible
enhance
sustainability
paradigm
direction
energy
security
via
usage
agroforestry
waste.
Scientific Reports,
Journal Year:
2023,
Volume and Issue:
13(1)
Published: April 13, 2023
Abstract
A
reliable
and
accurate
diagnosis
identification
system
is
required
to
prevent
manage
tea
leaf
diseases.
Tea
diseases
are
detected
manually,
increasing
time
affecting
yield
quality
productivity.
This
study
aims
present
an
artificial
intelligence-based
solution
the
problem
of
disease
detection
by
training
fastest
single-stage
object
model,
YOLOv7,
on
diseased
dataset
collected
from
four
prominent
gardens
in
Bangladesh.
4000
digital
images
five
types
these
gardens,
generating
a
manually
annotated,
data-augmented
image
dataset.
incorporates
data
augmentation
approaches
solve
issue
insufficient
sample
sizes.
The
results
for
YOLOv7
approach
validated
statistical
metrics
like
accuracy,
precision,
recall,
mAP
value,
F1-score,
which
resulted
97.3%,
96.7%,
96.4%,
98.2%,
0.965,
respectively.
Experimental
demonstrate
that
natural
scene
superior
existing
target
networks,
including
CNN,
Deep
DNN,
AX-Retina
Net,
improved
DCNN,
YOLOv5,
Multi-objective
segmentation.
Hence,
this
expected
minimize
workload
entomologists
aid
rapid
diseases,
thus
minimizing
economic
losses.
