B Cell–Inspired Algorithms - Immunological Computation: Theory and Applications - page 117

Chapter 5

B Cell-Inspired Algorithms

h is chapter describes clonal selection algorithms and artifi cial immune networks
(AINs), which are mainly inspired by B cells' response to antigens. First, the
main features of clonal selection algorithms and their similarities to evolutionary
approaches are presented. h en, continuous and discrete immune network (IN)
models are discussed. Finally, diff erent versions of IN model are described briefl y.

5.1 Clonal Selection Algorithms

Clonal selection algorithms are developed based on the clonal selection theory
(Burnet, 1959) proposed nearly 50 years ago. h e main immunological elements
used are

Maintenance of a specifi c memory set
Selection and cloning of most stimulated antibodies
Removal of poorly stimulated or nonstimulated antibodies
A nity maturation (hypermutation) of activated immune cells
Generation and maintenance of a diverse set of antibodies

Clonal selection algorithms (De Castro and Von Zuben, 2000), however, are very
similar to a kind of evolutionary algorithm; namely, evolutionary strategies (Beyer
and Schwefel, 2002), although they have a diff erent biological inspiration. Clonal
selection algorithms are also population-based search and optimization algorithms
generating a memory pool of suitable antibodies for solving a particular problem.

117