<s>
In	O
machine	O
learning	O
,	O
local	B-General_Concept
case-control	I-General_Concept
sampling	I-General_Concept
is	O
an	O
algorithm	O
used	O
to	O
reduce	O
the	O
complexity	O
of	O
training	O
a	O
logistic	O
regression	O
classifier	B-General_Concept
.	O
</s>
<s>
In	O
classification	B-General_Concept
,	O
a	O
dataset	O
is	O
a	O
set	O
of	O
N	O
data	O
points	O
,	O
where	O
is	O
a	O
feature	O
vector	O
,	O
is	O
a	O
label	O
.	O
</s>
<s>
Intuitively	O
these	O
samples	O
are	O
closer	O
to	O
the	O
decision	B-General_Concept
boundary	I-General_Concept
of	O
the	O
classifier	B-General_Concept
and	O
is	O
thus	O
more	O
informative	O
.	O
</s>
<s>
For	O
example	O
,	O
if	O
the	O
objective	O
is	O
to	O
use	O
a	O
subsample	O
with	O
size	O
,	O
first	O
estimate	O
a	O
model	O
using	O
samples	O
from	O
weighted	O
case	O
control	O
sampling	O
,	O
then	O
collect	O
another	O
samples	O
using	O
local	B-General_Concept
case-control	I-General_Concept
sampling	I-General_Concept
.	O
</s>
<s>
In	O
cases	O
where	O
the	O
number	O
of	O
samples	O
desired	O
is	O
precise	O
,	O
a	O
convenient	O
alternative	O
method	O
is	O
to	O
uniformly	O
downsample	O
from	O
a	O
larger	O
subsample	O
selected	O
by	O
local	B-General_Concept
case-control	I-General_Concept
sampling	I-General_Concept
.	O
</s>
<s>
When	O
the	O
pilot	O
is	O
consistent	O
,	O
the	O
estimates	O
using	O
the	O
samples	O
from	O
local	B-General_Concept
case-control	I-General_Concept
sampling	I-General_Concept
is	O
consistent	O
even	O
under	O
model	O
misspecification	O
.	O
</s>
