Information Technology Reference
In-Depth Information
Fig. 1
Amygdala model of Brain Emotional Learning (BELBIC structure)
there is a node for each sensory input. There is another node in Amygdala for Thala-
mus input, the amount of it is equal to maximum of sensory inputs. Output of nodes
in Amygdala and Orbitofrontal is calculated according to equations (1) and (2).
A
i
=
S
i
×
V
i
(1)
=
×
O
i
S
i
W
i
(2)
Where
A
i
and
O
i
are outputs of nodes in network structure of Amygdala and Or-
bitofrontal respectively,
V
and
W
are weights of nodes and
S
i
is sensory input.
Changes in
V
and
W
in learning process are calculated by equations (3) and (4)
respectively.
−
∑
j
Δ
V
i
=
α
×
max
(
0
,
S
i
×
(
R
A
j
))
(3)
×
(
∑
j
Δ
W
i
=
β
×
S
i
O
j
−
R
)
(4)
As it is shown, amount of
A
i
cannot be reduced which means that forgetting learned
information wont take place in Amygdala. In fact forgetting or lets say prevention
is Orbitofrontal duty. Finally model output obtains from equation (5).
=
∑
i
−
∑
i
E
A
i
O
i
(5)
Search WWH ::
Custom Search