<s>
Averaged	B-General_Concept
one-dependence	I-General_Concept
estimators	I-General_Concept
(	O
AODE	B-General_Concept
)	O
is	O
a	O
probabilistic	O
classification	B-General_Concept
learning	I-General_Concept
technique	O
.	O
</s>
<s>
It	O
was	O
developed	O
to	O
address	O
the	O
attribute-independence	O
problem	O
of	O
the	O
popular	O
naive	B-General_Concept
Bayes	I-General_Concept
classifier	I-General_Concept
.	O
</s>
<s>
It	O
frequently	O
develops	O
substantially	O
more	O
accurate	O
classifiers	B-General_Concept
than	O
naive	B-General_Concept
Bayes	I-General_Concept
at	O
the	O
cost	O
of	O
a	O
modest	O
increase	O
in	O
the	O
amount	O
of	O
computation	O
.	O
</s>
<s>
AODE	B-General_Concept
seeks	O
to	O
estimate	O
the	O
probability	O
of	O
each	O
class	O
y	O
given	O
a	O
specified	O
set	O
of	O
features	O
x1	O
,	O
...	O
xn	O
,	O
P( y	O
|	O
x1	O
,	O
...	O
xn	O
)	O
.	O
</s>
<s>
This	O
formula	O
defines	O
a	O
special	O
form	O
of	O
One	O
Dependence	O
Estimator	O
(	O
ODE	O
)	O
,	O
a	O
variant	O
of	O
the	O
naive	B-General_Concept
Bayes	I-General_Concept
classifier	I-General_Concept
that	O
makes	O
the	O
above	O
independence	O
assumption	O
that	O
is	O
weaker	O
(	O
and	O
hence	O
potentially	O
less	O
harmful	O
)	O
than	O
the	O
naive	B-General_Concept
Bayes	I-General_Concept
 '	O
independence	O
assumption	O
.	O
</s>
<s>
In	O
consequence	O
,	O
each	O
ODE	O
should	O
create	O
a	O
less	O
biased	O
estimator	O
than	O
naive	B-General_Concept
Bayes	I-General_Concept
.	O
</s>
<s>
AODE	B-General_Concept
reduces	O
this	O
variance	O
by	O
averaging	O
the	O
estimates	O
of	O
all	O
such	O
ODEs	O
.	O
</s>
<s>
Like	O
naive	B-General_Concept
Bayes	I-General_Concept
,	O
AODE	B-General_Concept
does	O
not	O
perform	O
model	O
selection	O
and	O
does	O
not	O
use	O
tuneable	O
parameters	O
.	O
</s>
<s>
It	O
supports	O
incremental	B-Algorithm
learning	I-Algorithm
whereby	O
the	O
classifier	B-General_Concept
can	O
be	O
updated	O
efficiently	O
with	O
information	O
from	O
new	O
examples	O
as	O
they	O
become	O
available	O
.	O
</s>
<s>
AODE	B-General_Concept
has	O
computational	O
complexity	O
at	O
training	O
time	O
and	O
at	O
classification	O
time	O
,	O
where	O
n	O
is	O
the	O
number	O
of	O
features	O
,	O
l	O
is	O
the	O
number	O
of	O
training	O
examples	O
and	O
k	O
is	O
the	O
number	O
of	O
classes	O
.	O
</s>
<s>
The	O
free	O
Weka	B-Language
machine	O
learning	O
suite	O
includes	O
an	O
implementation	O
of	O
AODE	B-General_Concept
.	O
</s>
