<s>
Predictive	B-General_Concept
analytics	I-General_Concept
is	O
a	O
form	O
of	O
business	B-Algorithm
analytics	I-Algorithm
applying	O
machine	O
learning	O
to	O
generate	O
a	O
predictive	B-General_Concept
model	I-General_Concept
for	O
certain	O
business	O
applications	O
.	O
</s>
<s>
As	O
such	O
,	O
it	O
encompasses	O
a	O
variety	O
of	O
statistical	O
techniques	O
from	O
predictive	B-General_Concept
modeling	I-General_Concept
and	O
machine	O
learning	O
that	O
analyze	O
current	O
and	O
historical	O
facts	O
to	O
make	O
predictions	O
about	O
future	O
or	O
otherwise	O
unknown	O
events	O
.	O
</s>
<s>
In	O
business	O
,	O
predictive	B-General_Concept
models	I-General_Concept
exploit	O
patterns	O
found	O
in	O
historical	O
and	O
transactional	O
data	O
to	O
identify	O
risks	O
and	O
opportunities	O
.	O
</s>
<s>
The	O
defining	O
functional	O
effect	O
of	O
these	O
technical	O
approaches	O
is	O
that	O
predictive	B-General_Concept
analytics	I-General_Concept
provides	O
a	O
predictive	O
score	O
(	O
probability	O
)	O
for	O
each	O
individual	O
(	O
customer	O
,	O
employee	O
,	O
healthcare	O
patient	O
,	O
product	O
SKU	O
,	O
vehicle	O
,	O
component	O
,	O
machine	O
,	O
or	O
other	O
organizational	O
unit	O
)	O
in	O
order	O
to	O
determine	O
,	O
inform	O
,	O
or	O
influence	O
organizational	O
processes	O
that	O
pertain	O
across	O
large	O
numbers	O
of	O
individuals	O
,	O
such	O
as	O
in	O
marketing	O
,	O
credit	O
risk	O
assessment	O
,	O
fraud	O
detection	O
,	O
manufacturing	O
,	O
healthcare	O
,	O
and	O
government	O
operations	O
including	O
law	O
enforcement	O
.	O
</s>
<s>
Predictive	B-General_Concept
analytics	I-General_Concept
is	O
a	O
set	O
of	O
business	B-General_Concept
intelligence	I-General_Concept
(	O
BI	O
)	O
technologies	O
that	O
uncovers	O
relationships	O
and	O
patterns	O
within	O
large	O
volumes	O
of	O
data	O
that	O
can	O
be	O
used	O
to	O
predict	O
behavior	O
and	O
events	O
.	O
</s>
<s>
Unlike	O
other	O
BI	O
technologies	O
,	O
predictive	B-General_Concept
analytics	I-General_Concept
is	O
forward-looking	O
,	O
using	O
past	O
events	O
to	O
anticipate	O
the	O
future	O
.	O
</s>
<s>
Predictive	B-General_Concept
analytics	I-General_Concept
statistical	O
techniques	O
include	O
data	B-Application
modeling	I-Application
,	O
machine	O
learning	O
,	O
AI	B-Application
,	O
deep	B-Algorithm
learning	I-Algorithm
algorithms	O
and	O
data	B-Application
mining	I-Application
.	O
</s>
<s>
Often	O
the	O
unknown	O
event	O
of	O
interest	O
is	O
in	O
the	O
future	O
,	O
but	O
predictive	B-General_Concept
analytics	I-General_Concept
can	O
be	O
applied	O
to	O
any	O
type	O
of	O
unknown	O
whether	O
it	O
be	O
in	O
the	O
past	O
,	O
present	O
or	O
future	O
.	O
</s>
<s>
The	O
core	O
of	O
predictive	B-General_Concept
analytics	I-General_Concept
relies	O
on	O
capturing	O
relationships	O
between	O
explanatory	O
variables	O
and	O
the	O
predicted	O
variables	O
from	O
past	O
occurrences	O
,	O
and	O
exploiting	O
them	O
to	O
predict	O
the	O
unknown	O
outcome	O
.	O
</s>
<s>
Predictive	B-General_Concept
analytics	I-General_Concept
is	O
often	O
defined	O
as	O
predicting	O
at	O
a	O
more	O
detailed	O
level	O
of	O
granularity	O
,	O
i.e.	O
,	O
generating	O
predictive	O
scores	O
(	O
probabilities	O
)	O
for	O
each	O
individual	O
organizational	O
element	O
.	O
</s>
<s>
For	O
example	O
,	O
"	O
Predictive	B-General_Concept
analytics	I-General_Concept
—	O
Technology	O
that	O
learns	O
from	O
experience	O
(	O
data	O
)	O
to	O
predict	O
the	O
future	O
behavior	O
of	O
individuals	O
in	O
order	O
to	O
drive	O
better	O
decisions.	O
"	O
</s>
<s>
In	O
future	O
industrial	O
systems	O
,	O
the	O
value	O
of	O
predictive	B-General_Concept
analytics	I-General_Concept
will	O
be	O
to	O
predict	O
and	O
prevent	O
potential	O
issues	O
to	O
achieve	O
near-zero	O
break-down	O
and	O
further	O
be	O
integrated	O
into	O
prescriptive	O
analytics	O
for	O
decision	O
optimization	O
.	O
</s>
<s>
The	O
approaches	O
and	O
techniques	O
used	O
to	O
conduct	O
predictive	B-General_Concept
analytics	I-General_Concept
can	O
broadly	O
be	O
grouped	O
into	O
regression	O
techniques	O
and	O
machine	O
learning	O
techniques	O
.	O
</s>
<s>
This	O
is	O
accomplished	O
through	O
artificial	B-Application
intelligence	I-Application
,	O
algorithms	O
,	O
and	O
models	O
.	O
</s>
<s>
Predictive	B-General_Concept
Modeling	I-General_Concept
is	O
a	O
statistical	O
technique	O
used	O
to	O
predict	O
future	O
behavior	O
.	O
</s>
<s>
It	O
utilizes	O
predictive	B-General_Concept
models	I-General_Concept
to	O
analyze	O
a	O
relationship	O
between	O
a	O
specific	O
unit	O
in	O
a	O
given	O
sample	O
and	O
one	O
or	O
more	O
features	O
of	O
the	O
unit	O
.	O
</s>
<s>
Predictive	B-General_Concept
model	I-General_Concept
solutions	O
can	O
be	O
considered	O
a	O
type	O
of	O
data	B-Application
mining	I-Application
technology	O
.	O
</s>
<s>
Regardless	O
of	O
the	O
methodology	O
used	O
,	O
in	O
general	O
,	O
the	O
process	O
of	O
creating	O
predictive	B-General_Concept
models	I-General_Concept
involves	O
the	O
same	O
steps	O
.	O
</s>
<s>
Auditors	O
accomplish	O
this	O
process	O
through	O
predictive	B-General_Concept
modeling	I-General_Concept
to	O
form	O
predictions	O
called	O
conditional	O
expectations	O
of	O
the	O
balances	O
being	O
audited	O
using	O
autoregressive	O
integrated	O
moving	O
average	O
(	O
ARIMA	O
)	O
methods	O
and	O
general	O
regression	O
analysis	O
methods	O
,	O
specifically	O
through	O
the	O
Statistical	O
Technique	O
for	O
Analytical	O
Review	O
(	O
STAR	O
)	O
methods	O
.	O
</s>
<s>
Predictive	B-General_Concept
analytics	I-General_Concept
can	O
be	O
used	O
and	O
is	O
capable	O
of	O
providing	O
many	O
benefits	O
to	O
a	O
wide	O
range	O
of	O
businesses	O
,	O
including	O
asset	O
management	O
firms	O
,	O
insurance	O
companies	O
,	O
communication	O
companies	O
,	O
and	O
many	O
other	O
firms	O
.	O
</s>
<s>
In	O
a	O
study	O
conducted	O
by	O
IDC	O
Analyze	O
the	O
Future	O
,	O
Dan	O
Vasset	O
and	O
Henry	O
D	O
.	O
Morris	O
explain	O
how	O
an	O
asset	O
management	O
firm	O
used	O
predictive	B-General_Concept
analytics	I-General_Concept
to	O
develop	O
a	O
better	O
marketing	O
campaign	O
.	O
</s>
<s>
Predictive	B-General_Concept
analytics	I-General_Concept
was	O
used	O
to	O
predict	O
the	O
likelihood	O
that	O
a	O
possible	O
customer	O
would	O
accept	O
a	O
personalized	O
offer	O
.	O
</s>
<s>
Due	O
to	O
the	O
marketing	O
campaign	O
and	O
predictive	B-General_Concept
analytics	I-General_Concept
,	O
the	O
firm	O
's	O
acceptance	O
rate	O
skyrocketed	O
,	O
with	O
three	O
times	O
the	O
number	O
of	O
people	O
accepting	O
their	O
personalized	O
offers	O
.	O
</s>
<s>
Technological	O
advances	O
in	O
predictive	B-General_Concept
analytics	I-General_Concept
have	O
increased	O
its	O
value	O
to	O
firms	O
.	O
</s>
<s>
One	O
technological	O
advancement	O
is	O
more	O
powerful	O
computers	O
,	O
and	O
with	O
this	O
predictive	B-General_Concept
analytics	I-General_Concept
has	O
become	O
able	O
to	O
create	O
forecasts	O
on	O
large	O
data	O
sets	O
much	O
faster	O
.	O
</s>
<s>
With	O
the	O
increased	O
computing	O
power	O
also	O
comes	O
more	O
data	O
and	O
applications	O
,	O
meaning	O
a	O
wider	O
array	O
of	O
inputs	O
to	O
use	O
with	O
predictive	B-General_Concept
analytics	I-General_Concept
.	O
</s>
<s>
Due	O
to	O
these	O
advancements	O
,	O
many	O
more	O
corporations	O
are	O
adopting	O
predictive	B-General_Concept
analytics	I-General_Concept
and	O
seeing	O
the	O
benefits	O
in	O
employee	O
efficiency	O
and	O
effectiveness	O
,	O
as	O
well	O
as	O
profits	O
.	O
</s>
<s>
Some	O
child	O
welfare	O
agencies	O
have	O
started	O
using	O
predictive	B-General_Concept
analytics	I-General_Concept
to	O
flag	O
high	O
risk	O
cases	O
.	O
</s>
<s>
For	O
example	O
,	O
in	O
Hillsborough	O
County	O
,	O
Florida	O
,	O
the	O
child	O
welfare	O
agency	O
's	O
use	O
of	O
a	O
predictive	B-General_Concept
modeling	I-General_Concept
tool	O
has	O
prevented	O
abuse-related	O
child	O
deaths	O
in	O
the	O
target	O
population	O
.	O
</s>
<s>
The	O
predicting	O
of	O
the	O
outcome	O
of	O
juridical	O
decisions	O
can	O
be	O
done	O
by	O
AI	B-Application
programs	O
.	O
</s>
<s>
These	O
types	O
of	O
problems	O
can	O
be	O
addressed	O
by	O
predictive	B-General_Concept
analytics	I-General_Concept
using	O
time	O
series	O
techniques	O
(	O
see	O
below	O
)	O
.	O
</s>
<s>
Predictive	B-General_Concept
analytics	I-General_Concept
can	O
help	O
underwrite	O
these	O
quantities	O
by	O
predicting	O
the	O
chances	O
of	O
illness	O
,	O
default	O
,	O
bankruptcy	O
,	O
etc	O
.	O
</s>
<s>
Predictive	B-General_Concept
analytics	I-General_Concept
can	O
streamline	O
the	O
process	O
of	O
customer	O
acquisition	O
by	O
predicting	O
the	O
future	O
risk	O
behavior	O
of	O
a	O
customer	O
using	O
application	O
level	O
data	O
.	O
</s>
<s>
Predictive	B-General_Concept
analytics	I-General_Concept
in	O
the	O
form	O
of	O
credit	O
scores	O
have	O
reduced	O
the	O
amount	O
of	O
time	O
it	O
takes	O
for	O
loan	O
approvals	O
,	O
especially	O
in	O
the	O
mortgage	O
market	O
.	O
</s>
<s>
Proper	O
predictive	B-General_Concept
analytics	I-General_Concept
can	O
lead	O
to	O
proper	O
pricing	O
decisions	O
,	O
which	O
can	O
help	O
mitigate	O
future	O
risk	O
of	O
default	O
.	O
</s>
<s>
Predictive	B-General_Concept
analytics	I-General_Concept
can	O
be	O
used	O
to	O
mitigate	O
moral	O
hazard	O
and	O
prevent	O
accidents	O
from	O
occurring	O
.	O
</s>
