Information Technology Reference
In-Depth Information
AiNet ()
1. Initialization
Create an initial network of antibodies
2. Antigenic presentation
While
not stopping criterion met
do
for
each antigenic pattern
do
2.1.
compute affinities with current antigen
for
each antibody
do
compute affinity with current antigen
2.2.
clonal selection
select a number of elements with high affinity and
reproduce (clone) them proportionally to their affinity
2.3.
metadynamics
remove memory clones whose affinity with current
antigen is less than a predefined threshold
2.4.
clonal interactions
determine the network interactions (affinity) among
elements in the clonal memory set
2.5.
clonal suppression
remove memory clones whose affinity with each other
is less than a pre-specified threshold
2.6.
network update
incorporate remaining clones of the clonal memory
with all network antibodies
3. Compute antibody interactions
o
network stimulations
compute similarity between each pair of network antibodies
o
network suppressions
eliminate all network antibodies whose affinity is less
than a pre-specified threshold
o
diversity handling
introduce a set of new randomly generated antibodies
into the network
Figure 5.7
The AiNet algorithm.
Search WWH ::
Custom Search