Information Technology Reference
In-Depth Information
1. Generate at random a new antibody cell (x,y) and add with concentration
10.
2. (Possibly) add a new antigen with coordinates ( x a ,y a ) and concentration
1000.
3. Calculate the stimulation S Ab of each antibody cell
4. If L<S Ab <U , increase the concentration of the cell by 1, otherwise
decrease it by 1
5. Calculate the stimulation S Ag of each antigen cell
6. If L<S Ag , decrease the concentration of the antigen according to S Ag / ( L
100).
7. Remove any cells whose concentration has reached 0.
Stimulation of cells and antigens is calculated according to the equations be-
low. For the complementary model, then Ab and Ag represent the complemen-
tary position of an antibody Ab or an antigen Ag , given by (100
y )for
a grid of dimension 100x100. For the similarity model, then Ab = Ab and Ag =
Ag .Theterms A c and E c represent the concentration of the antibody A and the
antigen E respectively, and r represents the recognition radius of the cell and
assumes a circular recognition region surrounding each complementary point.
x, 100
)+
cells E
Ab ||
Ab ||
S Ab =
A c ( r
−||
A
E c ( r
−||
E
)
( )
antigens A
Ag ||
S Ag =
E c ( r
−||
E
)
(2)
cells E
Binary Model. Except for the definition of clone identity, the binary model
closely follows the previous description regarding the 2D-shape space. Instead of
a point in a plane, each cell is now identified by a binary bit-string of N bits and
the anity a cell i exerts on another cell j is defined by the following equation
Affinity ( i, j ) = 100 .C i . ( HD ( i, j )
T ) / ( N bits
T )
(3)
with C i being the concentration of the cell i , HD the hamming distance between
the two bit-strings and T , the anity threshold, playing an equivalent role of the
parameter r in the 2D shape-space model. Like before, the total anity (field)
received by a cell i , S i , is obtained by summing the anity for all cells present
in the system, given that this anity can either be positive or null. Note that
antibodies can be stimulated by antigens and antibodies, while antigens can only
be stimulated by antibodies.
Keeping the system as resembling as possible to the 2D shape-space model,
the algorithm is as follows:
1. Generate at random a new antibody cell (bit-string) having an anity field
between L and U, with concentration 75.
2. (Possibly) add a new antigen with concentration 100.
3. Calculate the stimulation S Ab received by each antibody
Search WWH ::




Custom Search