The Endogenous Double Plasticity of the Immune Network and the Inspiration to be drawn for Engineering Artifacts

Although for reasons that are discussed, immunology have had a weak influence so far on the Ensign of artifacts, one key aspects of immune networks, namely their endogenous double plasticity could be of interest for future engineering applications facing complex, hard to Model and time-varying environments. In immune networks, this double plasticity allows the system to conduct its selfassertion role while being in constant shifting according to the organic"s ontogenic changes and in response to the environmental coupling. We argue that the Envelopment of complex systems, adaptive both at a structural and parametric levels, should comply with the following basic operational principles: - the structural adjustments intermittently occur following a longer time scale than the parametric adjustment - the structural plasticity amounts to the addition of new elements in the system and the suppression of redundant elements from it - the structural adjustments are EnpenEnnt on the temporal evolution of the internal parameters being subject to some learning i.e. when and how to perform a structural change should Enpend on data related to the dynamics of the parametric change so that the network endogenous behaviour and no exogenous criteria will guiEn these structural changes - these structural alterations have to be done in a “collective” spirit namely by applying heurisitics like “help the weakest elements (or compensate for them)”, “maintain diversity”, “fill the blank spaces”, “suppress redundant elements”. Three applications are briefly discussed: neural network classification, autonomous agent learning by reinforcement learnig and the control of chaos, where these principles are in full play and seem to have positive effect.

[1]  Hugues Bersini,et al.  The Immune Learning Mechanisms: Recruitment Reinforcement and their applications , 1993 .

[2]  F. Varela,et al.  Evidence for a functional idiotypic network among natural antibodies in normal mice. , 1989, Proceedings of the National Academy of Sciences of the United States of America.

[3]  Piero Mussio,et al.  Toward a Practice of Autonomous Systems , 1994 .

[4]  H. Bersini,et al.  An algorithm for self-structuring neural net classifiers , 1994, Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94).

[5]  F. Varela,et al.  Second generation immune networks. , 1991, Immunology today.

[6]  Richard S. Sutton,et al.  Reinforcement learning architectures for animats , 1991 .

[7]  D.E. Goldberg,et al.  Classifier Systems and Genetic Algorithms , 1989, Artif. Intell..

[8]  John H. Holland,et al.  Induction: Processes of Inference, Learning, and Discovery , 1987, IEEE Expert.

[9]  Hugues Bersini,et al.  Hints for Adaptive Problem Solving Gleaned from Immune Networks , 1990, PPSN.

[10]  Anne Lohrli.  Chapman and Hall , 1985 .

[11]  A. Perelson,et al.  Size and connectivity as emergent properties of a developing immune network. , 1991, Journal of theoretical biology.

[12]  John N. Tsitsiklis,et al.  Asynchronous stochastic approximation and Q-learning , 1993, Proceedings of 32nd IEEE Conference on Decision and Control.

[13]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[14]  Jeffrey O. Kephart,et al.  A biologically inspired immune system for computers , 1994 .

[15]  Yoshio Hirose,et al.  Backpropagation algorithm which varies the number of hidden units , 1989, International 1989 Joint Conference on Neural Networks.

[16]  Hugues Bersini.  REINFORCEMENT LEARNING AND RECRUITMENT MECHANISM FOR ADAPTIVE DISTRIBUTED CONTROL , 1993 .

[17]  Stephanie Forrest,et al.  Analogies with immunology represent an important step toward the vision of robust, distributed protection for computers. , 1991 .

[18]  Hugues Bersini,et al.  Immune Network and Adaptive Control , 1991 .

[19]  A S Perelson,et al.  Localized memories in idiotypic networks. , 1990, Journal of theoretical biology.

[20]  Ying-Cheng Lai,et al.  Controlling chaos , 1994 .

[21]  F. Varela,et al.  Development of an idiotypic network in shape space. , 1994, Journal of theoretical biology.