►
From YouTube: Time Series Forecasting in ML.NET
Description
How to use ML.NET to create predictions, or forecasts, on your time series data.
Code - https://github.com/jwood803/MLNetExamples/blob/master/MLNetExamples/TimeSeriesForecast/Program.cs
Detect anomalies in time series data in ML.NET - https://www.youtube.com/watch?v=Nu9lz5jNkGs
Detect anomalies with Cognitive Services - https://www.youtube.com/watch?v=gfb63wvjnYQ
ML.NET Playlist - https://www.youtube.com/watch?v=8gVhJKszzzI&list=PLl_upHIj19Zy3o09oICOutbNfXj332czx
Contact:
Twitter: https://twitter.com/JWood/
Blog: https://jonwood.co/
Gear used (affiliate links):
Mic - https://amzn.to/2YEXtxI
Mouse - https://amzn.to/2ZtASoQ
A
Hey
everyone
so
before
I've
done
a
video
on
Tom
series
data
where
it
can
detect
anomalies
in
your
data.
But
what,
if
you
wanted
to
make
predictions
well
in
this
video
I'll
show
how
you
can
create
an
M&M
model
that
can
make
predictions
or
forecasts
from
your
time
series
data,
so
I'm
here
in
visual
studio
and
get
a
done
that
or
console
project
and
I
have
my
data
in
here.
A
So
it's
just
two
columns,
first
as
to
date,
and
it's
an
early
date
here,
so
it
goes
from
each
our
own
of
for
each
day
and
then
the
energy
level
for
that
hour
is
also
enable
and
we're
gonna
build
a
model
that
predicts
the
next
four
energy
levels
from
this
data,
all
right
so
to
get
started.
Let's
add
some
nougat
packages
here.
First,
we'll
add
in
Monette.
A
Now
just
go
ahead
and
use
the
1.5
preview
version
here
and
since
we're
dealing
with
time
series
data,
we
need
to
add
a
separate
time
series
nougat
project,
it's
the
same
1.5
preview
version.
So
first
things.
First,
let's
create
our
context
and
then
we'll
load,
the
data
using
context,
data
load
from
text
file
and
I'll
load
it
into
an
energy
data
class
that
we
need
to
create.
But
I'll
do
that
in
just
a
moment
here.
A
A
So
this
class
only
has
two
items:
I'm
gonna
use
the
load
column
attribute
first
here
so
knows
where,
in
the
fall
to
pick
up
these
items,
first,
it's
going
to
be
that
date
and
I'm
gonna,
set
it
to
a
date/time
type.
You
know
just
call
it
dates,
and
the
next
one
is
gonna,
be
the
next
column
and
that's
that
floats
and
call
it
energy
all
right
with
that
done.
We
can
create
our
pipeline
and
I'll
do
context
that
forecasting
that
forecast
by
SSA
it's
gonna,
take
in
a
few
parameters
here.
A
The
first
one
is
gonna,
be
the
output
column
name.
This
is
what
we're
gonna
predict
on
here
and
so
I'll
just
do
forecast.
That's
why
I'm
gonna
put
as
my
prediction
name
here
and
the
next
one's
gonna
be
the
input
column,
and
so
with
that,
since
we
already
have
this
two
energy
data
item
here,
I'm
gonna
put
energy
items
as
my
input
and
next
we
have
a
few
other
parameters
here
and
we
can
just
we
could
play
around
with
these.
A
But
since
it's
a
forecast,
it's
going
to
be
a
forecasting
engine
and
to
do
that,
we
call
model
that
create
series
engine.
That's
gonna,
be
that
time
series
you
get
package
that
we
downloaded
it's
gonna,
taking
the
energy
data
as
the
inputs
ki-moon,
and
we
need
to
create
an
energy
forecast,
as
our
output
schema
not
parameter.
We
just
give
it
our
email
context,
all
right.
So,
let's
create
our
forecast
file
here.
A
This
is
going
to
be
one
property
here,
floats
an
array
of
floats
and
now
to
call
it
forecast
and
say
that
we
have
our
forecasting
engine
now
can
get
our
forecasts
on
it.
Movie
forecast
Gina
predict
with
this
harassment
remedies
how
many
forecasts
that
were
you
tell
it
to
to
get
and
if
we
want
to
change
that,
we
can
put
it
in
here
and
this
horizon
parameter
as
well.
If
we
want
to
change
that
from
what
we
put
in
our
pipeline,
so
we
have
our
new
forecasts,
so
we
have
a
forecast.
A
A
Just
run
this
and
see
what
we
get
here
there
yet
so
we
got
our
next
four
forecast
here
from
our
time
series
data
that
we
got
so,
as
you
can
see,
is
pretty
simple:
to
use
my
net
to
click,
your
time,
series
predictions
out
and
things
here
and
hope
you
enjoyed
the
video
and
I'll,
see
you
on
this
time.
Thanks.