Artificial Immune Systems

Abstract

The human immune system has numerous properties that make it ripe for exploitation in the computational domain, such as robustness and fault tolerance, and many different algorithms, collectively termed Artificial Immune Systems (AIS), have been inspired by it. Two generations of AIS are currently in use, with the first generation relying on simplified immune models and the second generation utilising interdisciplinary collaboration to develop a deeper understanding of the immune system and hence produce more complex models. Both generations of algorithms have been successfully applied to a variety of problems, including anomaly detection, pattern recognition, optimisation and robotics. In this chapter an overview of AIS is presented, its evolution is discussed, and it is shown that the diversification of the field is linked to the diversity of the immune system itself, leading to a number of algorithms as opposed to one archetypal system. Two case studies are also presented to help provide insight into the mechanisms of AIS; these are the idiotypic network approach and the Dendritic Cell Algorithm.

The Design and Development of Spam Risk Assessment Prototype: In Silico of Danger Theory Variants

“…The adaptation of biological ideas into the application of real world problem is involved a few phases as developed by Stepney in 2004 and known as Conceptual Framework [44] approach. This methodology employs an iterative approach to the creation and testing of novel immuneinspired algorithms and consists of 4 stages that are identified as:…”

Artificial Immune System Research in the past years-an overview of existing methods, new initiatives and applications in intrusion detection systems

“…The clonal selection mechanism [1,2,3]proposed by Burnet in 1959 and it describes the B-cell cloning process as shown in Fig.1,basically says that only those cells that can recognize the antigens can proliferate and hence get selected over the others.The important features of this theory are- Firstly the B-cells are selected to be fit for the purpose during a "training period".Those B-cells that exhibit cell-surface receptors(antibodies) that match the corresponding antigen are selected to form the initial B-cell population while the rest that cannot bind to the antigen are removed.The selected B-cells are then released to the periphery.When they encounter the antigen,multiple versions of the B-cell receptors(antibodies) are produced that can bind to the matching antigen.  The process of antibody tuning occurs through somatic hypermutation and affinity maturation [1].Somatic hypermutation ensures that the B-cell that exactly matches the antigen does not produce exact clones of itself that is the clones are slight variants of the parent cells.This is a kind of biological optimization that ultimately results in the production of antibodies that can bind more successfully to the antigens and hence may be used as detectors of non-self entities in the body.This whole process is called affinity maturation that produces the most responsive antibodies [1].…”

“… The process of antibody tuning occurs through somatic hypermutation and affinity maturation [1].Somatic hypermutation ensures that the B-cell that exactly matches the antigen does not produce exact clones of itself that is the clones are slight variants of the parent cells.This is a kind of biological optimization that ultimately results in the production of antibodies that can bind more successfully to the antigens and hence may be used as detectors of non-self entities in the body.This whole process is called affinity maturation that produces the most responsive antibodies [1].  When clonal selection operates on B-cells these can differentiate into long-lived memory cells [2] so as to encounter the repeated attacks by the same antigen during the entire lifetime of the individual.The initial exposure to an antigen induces an adaptive immune response that is initially handled by a small number of B-cells that produce antibodies of varying affinities.If we can store some high affinity antibody producing cells from the first infection so as to form a large initial specific B-cell sub-population(clone),then the effectiveness and speed of protection would be much more during later attacks [2].…”

“…The negative selection algorithm was proposed by Forrest et al in 1994 [3].It was inspired by the ability of the immune system to train the T-cells to recognize only the antigens(non-self) and to prevent them from detecting the body"s own cells(self).Basically we generate a set of (binary)detectors by randomly making candidates and then reject those that discard training self-data.A major application of a true NSA was in a system called "Lisys" [1] which had three steps- Detecting self(a set S of elements of length l in shape space)  Generating detectors(a set D of detectors such that no element of D matches any of the elements of S)  Monitoring the occurrence of anomalies/anomaly detection(monitor a continual data stream by continuously matching D against the data stream)…”

“…The Artificial Immune System(AIS),however exhibits a very high level of complexity.Different algorithms used in AIS actually imitate the behavior of the different types of cells of the immune system.In the first generation AIS,emphasis has been on Clonal Selection [1,2,3],Negative Selection [1,2,3] and Idiotypic approaches [1,2,3]which have often shown considerable limitations.So in recent years lot of developments have occurred and the second generation of AIS uses the Dendritic Cell Algorithm [4] and the Danger Theory [1,5]approach.AIS can certainly contribute with the already established methodologies in order to mutually improve their performances and application domains.…”