<s>
Data	B-General_Concept
binning	I-General_Concept
,	O
also	O
called	O
data	O
discrete	O
binning	O
or	O
data	O
bucketing	B-General_Concept
,	O
is	O
a	O
data	B-General_Concept
pre-processing	I-General_Concept
technique	O
used	O
to	O
reduce	O
the	O
effects	O
of	O
minor	O
observation	O
errors	O
.	O
</s>
<s>
The	O
original	O
data	O
values	O
which	O
fall	O
into	O
a	O
given	O
small	O
interval	O
,	O
a	O
bin	B-Data_Structure
,	O
are	O
replaced	O
by	O
a	O
value	O
representative	O
of	O
that	O
interval	O
,	O
often	O
a	O
central	O
value	O
(	O
mean	O
or	O
median	O
)	O
.	O
</s>
<s>
It	O
is	O
related	O
to	O
quantization	B-Algorithm
:	O
data	B-General_Concept
binning	I-General_Concept
operates	O
on	O
the	O
abscissa	O
axis	O
while	O
quantization	B-Algorithm
operates	O
on	O
the	O
ordinate	O
axis	O
.	O
</s>
<s>
Binning	O
is	O
a	O
generalization	O
of	O
rounding	B-Algorithm
.	O
</s>
<s>
Statistical	O
data	B-General_Concept
binning	I-General_Concept
is	O
a	O
way	O
to	O
group	O
numbers	O
of	O
more-or-less	O
continuous	O
values	O
into	O
a	O
smaller	O
number	O
of	O
"	O
bins	B-Data_Structure
"	O
.	O
</s>
<s>
It	O
can	O
also	O
be	O
used	O
in	O
multivariate	B-General_Concept
statistics	I-General_Concept
,	O
binning	O
in	O
several	O
dimensions	O
at	O
once	O
.	O
</s>
<s>
In	O
digital	B-Algorithm
image	I-Algorithm
processing	I-Algorithm
,	O
"	O
binning	O
"	O
has	O
a	O
very	O
different	O
meaning	O
.	O
</s>
<s>
Pixel	B-Algorithm
binning	I-Algorithm
is	O
the	O
process	O
of	O
combining	O
blocks	O
of	O
adjacent	O
pixels	B-Algorithm
throughout	O
an	O
image	O
,	O
by	O
summing	O
or	O
averaging	O
their	O
values	O
,	O
during	O
or	O
after	O
readout	O
.	O
</s>
<s>
Histograms	B-Algorithm
are	O
an	O
example	O
of	O
data	B-General_Concept
binning	I-General_Concept
used	O
in	O
order	O
to	O
observe	O
underlying	O
frequency	O
distributions	O
.	O
</s>
<s>
Data	B-General_Concept
binning	I-General_Concept
may	O
be	O
used	O
when	O
small	O
instrumental	O
shifts	O
in	O
the	O
spectral	O
dimension	O
from	O
mass	O
spectrometry	O
(	O
MS	O
)	O
or	O
nuclear	O
magnetic	O
resonance	O
(	O
NMR	O
)	O
experiments	O
will	O
be	O
falsely	O
interpreted	O
as	O
representing	O
different	O
components	O
,	O
when	O
a	O
collection	O
of	O
data	O
profiles	O
is	O
subjected	O
to	O
pattern	O
recognition	O
analysis	O
.	O
</s>
<s>
A	O
straightforward	O
way	O
to	O
cope	O
with	O
this	O
problem	O
is	O
by	O
using	O
binning	O
techniques	O
in	O
which	O
the	O
spectrum	O
is	O
reduced	O
in	O
resolution	O
to	O
a	O
sufficient	O
degree	O
to	O
ensure	O
that	O
a	O
given	O
peak	O
remains	O
in	O
its	O
bin	B-Data_Structure
despite	O
small	O
spectral	O
shifts	O
between	O
analyses	O
.	O
</s>
<s>
Also	O
,	O
several	O
digital	B-Device
camera	I-Device
systems	O
incorporate	O
an	O
automatic	O
pixel	B-Algorithm
binning	I-Algorithm
function	O
to	O
improve	O
image	O
contrast	O
.	O
</s>
<s>
Binning	O
is	O
also	O
used	O
in	O
machine	O
learning	O
to	O
speed	O
up	O
the	O
decision-tree	O
boosting	B-Algorithm
method	O
for	O
supervised	O
classification	O
and	O
regression	O
in	O
algorithms	O
such	O
as	O
Microsoft	O
's	O
LightGBM	B-Algorithm
and	O
scikit-learn	B-Application
'	O
s	O
.	O
</s>
