<s>
In	O
anomaly	B-Algorithm
detection	I-Algorithm
,	O
the	O
local	B-Algorithm
outlier	I-Algorithm
factor	I-Algorithm
(	O
LOF	O
)	O
is	O
an	O
algorithm	O
proposed	O
by	O
Markus	O
M	O
.	O
Breunig	O
,	O
Hans-Peter	O
Kriegel	O
,	O
Raymond	O
T	O
.	O
Ng	O
and	O
Jörg	O
Sander	O
in	O
2000	O
for	O
finding	O
anomalous	O
data	O
points	O
by	O
measuring	O
the	O
local	O
deviation	O
of	O
a	O
given	O
data	O
point	O
with	O
respect	O
to	O
its	O
neighbours	O
.	O
</s>
<s>
LOF	O
shares	O
some	O
concepts	O
with	O
DBSCAN	B-Algorithm
and	O
OPTICS	B-Algorithm
such	O
as	O
the	O
concepts	O
of	O
"	O
core	O
distance	O
"	O
and	O
"	O
reachability	O
distance	O
"	O
,	O
which	O
are	O
used	O
for	O
local	O
density	O
estimation	O
.	O
</s>
<s>
The	O
local	B-Algorithm
outlier	I-Algorithm
factor	I-Algorithm
is	O
based	O
on	O
a	O
concept	O
of	O
a	O
local	O
density	O
,	O
where	O
locality	O
is	O
given	O
by	O
k	O
nearest	O
neighbors	O
,	O
whose	O
distance	O
is	O
used	O
to	O
estimate	O
the	O
density	O
.	O
</s>
<s>
Objects	O
that	O
belong	O
to	O
the	O
k	O
nearest	O
neighbors	O
of	O
B	O
(	O
the	O
"	O
core	O
"	O
of	O
B	O
,	O
see	O
DBSCAN	B-Algorithm
cluster	I-Algorithm
analysis	I-Algorithm
)	O
are	O
considered	O
to	O
be	O
equally	O
distant	O
.	O
</s>
<s>
Feature	O
Bagging	O
for	O
Outlier	B-Algorithm
Detection	I-Algorithm
runs	O
LOF	O
on	O
multiple	O
projections	O
and	O
combines	O
the	O
results	O
for	O
improved	O
detection	O
qualities	O
in	O
high	O
dimensions	O
.	O
</s>
<s>
This	O
is	O
the	O
first	O
ensemble	B-Algorithm
learning	I-Algorithm
approach	O
to	O
outlier	B-Algorithm
detection	I-Algorithm
,	O
for	O
other	O
variants	O
see	O
ref	O
.	O
</s>
<s>
Interpreting	O
and	O
Unifying	O
Outlier	O
Scores	O
proposes	O
a	O
normalization	O
of	O
the	O
LOF	O
outlier	O
scores	O
to	O
the	O
interval	O
using	O
statistical	O
scaling	O
to	O
increase	O
usability	B-General_Concept
and	O
can	O
be	O
seen	O
an	O
improved	O
version	O
of	O
the	O
LoOP	O
ideas	O
.	O
</s>
<s>
On	O
Evaluation	O
of	O
Outlier	O
Rankings	O
and	O
Outlier	O
Scores	O
proposes	O
methods	O
for	O
measuring	O
similarity	O
and	O
diversity	O
of	O
methods	O
for	O
building	O
advanced	O
outlier	B-Algorithm
detection	I-Algorithm
ensembles	B-Algorithm
using	O
LOF	O
variants	O
and	O
other	O
algorithms	O
and	O
improving	O
on	O
the	O
Feature	O
Bagging	O
approach	O
discussed	O
above	O
.	O
</s>
<s>
Local	O
outlier	B-Algorithm
detection	I-Algorithm
reconsidered	O
:	O
a	O
generalized	O
view	O
on	O
locality	O
with	O
applications	O
to	O
spatial	O
,	O
video	O
,	O
and	O
network	O
outlier	B-Algorithm
detection	I-Algorithm
discusses	O
the	O
general	O
pattern	O
in	O
various	O
local	O
outlier	B-Algorithm
detection	I-Algorithm
methods	O
(	O
including	O
,	O
e.g.	O
,	O
LOF	O
,	O
a	O
simplified	O
version	O
of	O
LOF	O
and	O
LoOP	O
)	O
and	O
abstracts	O
from	O
this	O
into	O
a	O
general	O
framework	O
.	O
</s>
