<s>
Functional	B-Algorithm
Holography	I-Algorithm
(	O
FH	O
)	O
is	O
a	O
method	O
of	O
analysis	O
designed	O
to	O
extract	O
the	O
maximum	O
amount	O
of	O
functional	O
information	O
about	O
the	O
dynamical	O
network	O
as	O
a	O
whole	O
unit	O
.	O
</s>
<s>
Itay	O
Baruchi	O
and	O
his	O
Ph.D.	O
supervisor	O
,	O
Eshel	O
Ben-Jacob	O
,	O
introduced	O
the	O
Functional	B-Algorithm
Holography	I-Algorithm
(	O
FH	O
)	O
methodology	O
.	O
</s>
<s>
The	O
FH	O
analysis	O
was	O
devised	O
to	O
study	O
the	O
dynamics	O
of	O
task	O
performing	O
dynamical	O
networks	O
(	O
such	O
as	O
brain	O
activity	O
and	O
neural	B-Architecture
networks	I-Architecture
,	O
and	O
gene	O
networks	O
or	O
recorded	O
data	O
of	O
dynamics	O
system	O
such	O
as	O
stock	O
market	O
parameters	O
or	O
biological	O
chips	O
activities	O
)	O
.	O
</s>
<s>
The	O
Functional	B-Algorithm
Holography	I-Algorithm
analysis	O
method	O
was	O
first	O
introduced	O
in	O
2004	O
by	O
Itai	O
Baruchi	O
and	O
Eshel	O
Ben-Jacob	O
,	O
for	O
analysis	O
of	O
recorded	O
human	O
brain	O
activity	O
.	O
</s>
<s>
The	O
term	O
hologram	B-Algorithm
stands	O
for	O
“	O
whole	O
”	O
—	O
holo	O
in	O
Greek	O
,	O
plus	O
“	O
information	O
”	O
or	O
“	O
message	O
”	O
—	O
gram	O
in	O
Greek	O
.	O
</s>
<s>
In	O
a	O
holographic	B-Algorithm
photography	O
,	O
the	O
information	O
describing	O
a	O
3D	O
object	O
is	O
encoded	O
on	O
a	O
two-dimensional	O
photographic	O
film	O
,	O
ready	O
to	O
be	O
regenerated	O
into	O
a	O
holographic	B-Algorithm
image	I-Algorithm
or	O
hologram	B-Algorithm
.	O
</s>
<s>
Another	O
property	O
is	O
high	O
tolerance	O
to	O
noise	O
and	O
high	O
robustness	O
to	O
lesion	O
:	O
even	O
with	O
many	O
imperfections	O
or	O
with	O
several	O
pixels	O
removed	O
,	O
the	O
image	O
of	O
the	O
object	O
as	O
a	O
whole	O
is	O
still	O
retained	O
in	O
the	O
hologram	B-Algorithm
.	O
</s>
<s>
Another	O
related	O
feature	O
is	O
the	O
holographic	B-Algorithm
superposition	O
—	O
when	O
illuminated	O
together	O
(	O
placed	O
side	O
by	O
side	O
)	O
,	O
two	O
holograms	B-Algorithm
can	O
generate	O
a	O
superposition	O
of	O
the	O
corresponding	O
two	O
3D	O
objects	O
.	O
</s>
<s>
Superposition	O
of	O
objects	O
can	O
also	O
be	O
made	O
by	O
imprinting	O
the	O
images	O
of	O
the	O
two	O
(	O
or	O
more	O
)	O
3D	O
objects	O
on	O
the	O
same	O
holographic	B-Algorithm
film	O
.	O
</s>
<s>
These	O
and	O
other	O
special	O
features	O
of	O
hologram	B-Algorithm
are	O
due	O
to	O
the	O
way	O
the	O
information	O
is	O
encoded	O
on	O
the	O
films	O
—	O
not	O
a	O
direct	O
projection	O
of	O
the	O
picture	O
in	O
real	O
space	O
but	O
in	O
the	O
correlations	O
between	O
the	O
pixels	O
.	O
</s>
<s>
The	O
above	O
properties	O
of	O
holograms	B-Algorithm
guided	O
the	O
development	O
and	O
are	O
the	O
rationale	O
behind	O
the	O
functional	B-Algorithm
holography	I-Algorithm
method	O
presented	O
here	O
.	O
</s>
<s>
The	O
Functional	B-Algorithm
Holography	I-Algorithm
methodology	O
shares	O
the	O
special	O
features	O
of	O
holograms	B-Algorithm
—	O
tolerance	O
to	O
noise	O
,	O
robustness	O
to	O
lesion	O
,	O
holographic	B-Algorithm
superposition	O
and	O
holographic	B-Algorithm
zooming	O
.	O
</s>
<s>
The	O
Projection	O
of	O
affinity	O
matrix	O
using	O
dimension	O
reduction	O
algorithms	O
(	O
the	O
Principal	B-Application
Component	I-Application
Analysis	I-Application
,	O
PCA	O
)	O
onto	O
a	O
principal	O
three-dimensional	O
space	O
of	O
the	O
leading	O
eigenvectors	O
computed	O
by	O
the	O
algorithm	O
.	O
</s>
<s>
Retrieval	O
of	O
information	O
that	O
is	O
lost	O
in	O
the	O
dimension	O
reduction	O
-	O
the	O
nodes	O
are	O
connected	O
by	O
color-coded	O
lines	O
that	O
represent	O
the	O
level	O
of	O
similarities	O
,	O
which	O
is	O
then	O
used	O
to	O
construct	O
a	O
holographic	B-Algorithm
network	O
in	O
the	O
principal	O
space	O
.	O
</s>
