Information Technology Reference
In-Depth Information
GAIN (General Artificial Immune Network)
Input:
'
A
' as a set of antigens
Output:
Immune network consisting of '
B
' as a set of B-cells and connections between
them
1:
Initialization
1.1:
Assign
B
an initial set of B-cells
1.2:
Initialize network structure
L
2:
Repeat until
a stop criteria is met
2.1:
Antigen presentation:
Antigen / B-cell affinity
2.1.1:
Calculate
f
affinity
(
a
,
b
)
∀
a
∈
A
,
b
∈
B
Antigen/ B-cell stimulation
2.1.2: Calculate
f
stimulation
(
b
,
a
)
∀
a
∈
A
,
b
∈
B
2.2:
B-cell interaction:
B-cell /B-cell stimulation/suppression
Calculate
f
stimulation
(
b
,
b
) and
f
suppression
(
b
2.2.1:
′
′
,
b
)
∀
b
′
,
b
∈
B
2.3:
Affinity maturation:
Total stimulation
Σ
2.3.1: Calculate
F
(
b
):=
f
stimulation
(
a
,
b
) +
f
stimulation
(
b
′
,
b
) +
f
suppression
(
b
′
,
b
)
∈
A,b
′
∈
B,b
′
b
a
2.3.2:
Create
f
cloning
(
b
) clones to the B-cell
b
and mutate then
2.3.3:
Calculate stimulation of all new B-cells
2.4:
Metadynamics:
Detection/creation of B-cells and links
2.4.1:
Update network structure
L
Return immune network
3:
Return (
B
,
L
)
Figure 5.14 A GAIN algorithm. (From González, F., J. Galeano and A. Veloza, in
Proceedings of the 2005 Conference on Genetic and Evolutionary Computation
(GECCO'05),
ACM Press, Washington, 2005, 361-368.)
Search WWH ::
Custom Search