ABSTRACT
This chapter shows that the immune system affinity maturation via somatic mutation can be interpreted in terms of a learning mechanism of the second kind. It describes learning in simple neural networks, the so-called Hebb's rule and discusses qualitatively the possible existence of a similar rule for the immune system. The metaphors of cognition are often used when discussing the properties of the immune system. Attractors of the dynamics are interpreted in terms of memories that can be retrieved by inputting some "pattern" to a neural network or giving antigen to an immune network. In the limit of small source and mutation terms, the high steady-state populations characteristics of the immune or suppressed states are obtained when the proliferation function matches the death rate. In neural networks it is generally assumed that the pattern recognition process is faster than the learning process.
