<s>
Quickprop	B-Algorithm
is	O
an	O
iterative	O
method	O
for	O
determining	O
the	O
minimum	O
of	O
the	O
loss	O
function	O
of	O
an	O
artificial	B-Architecture
neural	I-Architecture
network	I-Architecture
,	O
following	O
an	O
algorithm	O
inspired	O
by	O
the	O
Newton	O
's	O
method	O
.	O
</s>
<s>
The	O
procedure	O
requires	O
only	O
local	O
information	O
of	O
the	O
artificial	B-Algorithm
neuron	I-Algorithm
to	O
which	O
it	O
is	O
applied	O
.	O
</s>
<s>
The	O
Quickprop	B-Algorithm
algorithm	O
is	O
an	O
implementation	O
of	O
the	O
error	O
backpropagation	B-Algorithm
algorithm	O
,	O
but	O
the	O
network	O
can	O
behave	O
chaotically	O
during	O
the	O
learning	O
phase	O
due	O
to	O
large	O
step	O
sizes	O
.	O
</s>
