<s>
In	O
machine	O
learning	O
,	O
a	O
Hyper	B-Algorithm
basis	I-Algorithm
function	I-Algorithm
network	I-Algorithm
,	O
or	O
HyperBF	O
network	O
,	O
is	O
a	O
generalization	O
of	O
radial	O
basis	O
function	O
(	O
RBF	O
)	O
networks	O
concept	O
,	O
where	O
the	O
Mahalanobis-like	O
distance	O
is	O
used	O
instead	O
of	O
Euclidean	O
distance	O
measure	O
.	O
</s>
<s>
Hyper	B-Algorithm
basis	I-Algorithm
function	I-Algorithm
networks	I-Algorithm
were	O
first	O
introduced	O
by	O
Poggio	O
and	O
Girosi	O
in	O
the	O
1990	O
paper	O
“	O
Networks	O
for	O
Approximation	O
and	O
Learning	O
”	O
.	O
</s>
<s>
The	O
typical	O
HyperBF	O
network	O
structure	O
consists	O
of	O
a	O
real	O
input	O
vector	O
,	O
a	O
hidden	O
layer	O
of	O
activation	B-Algorithm
functions	I-Algorithm
and	O
a	O
linear	O
output	O
layer	O
.	O
</s>
<s>
This	O
case	O
corresponds	O
to	O
the	O
regular	O
RBF	B-Algorithm
network	I-Algorithm
.	O
</s>
