Information Technology Reference
In-Depth Information
alg o rithm is presented next.
RAIN( )
1. Initialization
Create an initial network, select the initial set of B cells as a subset of the antigens
2. Antigenic presentation
While not stopping criterion met do
for each antigen do
2.1. compute network stimulation levels and clonal selection
for each B cell do
compute stimulation level
2.2. metadynamics
Remove B cells with low stimulation level, via the resource
allocation mechanism
2.3. clonal expansion
Select most stimulated B-cells and reproduce them in proportion to
their stimulation level
2.4. somatic hypermutation
Mutate each clone inversely proportional to its stimulation level
2.5. network update
select mutated clones to be incorporated in the network
Figure 5.6
Details of the RAIN algorithm.
are done by artifi cial cytokines, which are represented by a single clone of the
transmitting FRS. h e signal is received by a fractal receptor (a clone of the receiving
FRS) and then, the distance is calculated. If the distance is below a certain threshold
and FRS is mature, then the transmitting FRS is stimulated. h is stimulated FRS is
cloned with a fi xed probability and it is merged with the antigen by the merge process
of fractal proteins. If there exists no such FRS, a new one is created at the antigen
point as a primary response, like its parent model. At each iteration, based on the
stimulation level, the FRS concentration is increased. If this concentration is below a
mortality threshold, the element is removed from the network.
Search WWH ::




Custom Search