<s>
Split	B-Algorithm
Up	I-Algorithm
is	O
an	O
intelligent	O
decision	B-Application
support	I-Application
system	I-Application
,	O
which	O
makes	O
predictions	O
about	O
the	O
distribution	O
of	O
marital	O
property	O
following	O
divorce	O
in	O
Australia	O
.	O
</s>
<s>
Split	B-Algorithm
Up	I-Algorithm
operates	O
as	O
a	O
hybrid	O
system	O
,	O
combining	O
rule	O
–	O
based	O
reasoning	O
with	O
neural	B-Architecture
network	I-Architecture
theory	O
.	O
</s>
<s>
Neural	B-Architecture
networks	I-Architecture
,	O
by	O
contrast	O
,	O
are	O
considered	O
to	O
be	O
better	O
suited	O
to	O
generate	O
decisions	O
in	O
uncertain	O
domains	O
,	O
since	O
they	O
can	O
be	O
taught	O
to	O
weigh	O
the	O
factors	O
considered	O
by	O
judicial	O
decision	O
makers	O
from	O
case	O
data	O
.	O
</s>
<s>
Split_up	O
determines	O
the	O
percentage	O
split	O
by	O
using	O
a	O
combination	O
of	O
rule	O
-	O
based	O
reasoning	O
and	O
neural	B-Architecture
networks	I-Architecture
.	O
</s>
<s>
By	O
contrast	O
,	O
the	O
fourteen	O
dashed	O
arcs	O
establish	O
inferences	O
performed	O
with	O
the	O
use	O
of	O
neural	B-Architecture
networks	I-Architecture
.	O
</s>
<s>
In	O
Split_Up	O
,	O
the	O
neural	B-Architecture
network	I-Architecture
is	O
a	O
statistical	O
technique	O
for	O
learning	O
the	O
weights	O
of	O
each	O
of	O
the	O
relevant	O
attributes	O
used	O
in	O
a	O
percentage	O
split	O
determination	O
of	O
marital	O
property	O
.	O
</s>
<s>
Hence	O
the	O
inputs	O
to	O
the	O
neural	B-Architecture
network	I-Architecture
are	O
contributions	O
,	O
future	O
needs	O
and	O
wealth	O
,	O
and	O
the	O
output	O
the	O
percentage	O
split	O
predicted	O
.	O
</s>
<s>
Using	O
back	O
propagation	O
the	O
neural	B-Architecture
network	I-Architecture
learns	O
the	O
necessary	O
pattern	O
to	O
recognize	O
the	O
prediction	O
.	O
</s>
<s>
It	O
is	O
trained	O
by	O
repeatedly	O
exposing	O
it	O
to	O
examples	O
of	O
the	O
problem	O
and	O
learning	O
the	O
significance	O
(	O
weights	O
)	O
of	O
the	O
input	O
nodes	B-Algorithm
.	O
</s>
<s>
The	O
neural	B-Architecture
network	I-Architecture
used	O
by	O
Split_up	O
is	O
said	O
to	O
generalise	O
well	O
if	O
the	O
output	O
of	O
the	O
network	O
is	O
correct	O
(	O
or	O
nearly	O
correct	O
)	O
for	O
examples	O
not	O
seen	O
during	O
training	O
,	O
which	O
classifies	O
it	O
as	O
an	O
intelligent	B-Application
system	I-Application
.	O
</s>
<s>
Here	O
,	O
a	O
neural	B-Architecture
network	I-Architecture
(	O
or	O
rules	O
)	O
,	O
produce	O
a	O
conclusion	O
from	O
the	O
data	O
of	O
an	O
argument	O
and	O
the	O
data	O
,	O
warrant	O
and	O
backing	O
are	O
reproduced	O
to	O
generate	O
an	O
explanation	O
.	O
</s>
