<s>
In	O
statistics	O
,	O
several	O
scatterplot	B-Application
smoothing	I-Application
methods	O
are	O
available	O
to	O
fit	O
a	O
function	O
through	O
the	O
points	O
of	O
a	O
scatterplot	B-Application
to	O
best	O
represent	O
the	O
relationship	O
between	O
the	O
variables	O
.	O
</s>
<s>
Scatterplots	B-Application
may	O
be	O
smoothed	B-Application
by	O
fitting	O
a	O
line	O
to	O
the	O
data	O
points	O
in	O
a	O
diagram	O
.	O
</s>
<s>
This	O
line	O
attempts	O
to	O
display	O
the	O
non-random	O
component	O
of	O
the	O
association	O
between	O
the	O
variables	O
in	O
a	O
2D	O
scatter	B-Application
plot	I-Application
.	O
</s>
<s>
Smoothing	B-Application
attempts	O
to	O
separate	O
the	O
non-random	O
behaviour	O
in	O
the	O
data	O
from	O
the	O
random	O
fluctuations	O
,	O
removing	O
or	O
reducing	O
these	O
fluctuations	O
,	O
and	O
allows	O
prediction	O
of	O
the	O
response	O
based	O
value	O
of	O
the	O
explanatory	O
variable	O
.	O
</s>
<s>
Smoothing	B-Application
is	O
normally	O
accomplished	O
by	O
using	O
any	O
one	O
of	O
the	O
techniques	O
mentioned	O
below	O
.	O
</s>
<s>
The	O
smoothing	B-Application
curve	O
is	O
chosen	O
so	O
as	O
to	O
provide	O
the	O
best	O
fit	O
in	O
some	O
sense	O
,	O
often	O
defined	O
as	O
the	O
fit	O
that	O
results	O
in	O
the	O
minimum	O
sum	B-Algorithm
of	I-Algorithm
the	I-Algorithm
squared	I-Algorithm
errors	I-Algorithm
(	O
a	O
least	B-Algorithm
squares	I-Algorithm
criterion	O
)	O
.	O
</s>
