Memory in Idiotypic Network Dynamics
Francisco, Hani Louie (advisor: Klemm, Konstantin)
Master Thesis (2018)
A simple model for an idiotypic network is used to look at memory in the immune system. In the model, a network node corresponds to an idiotype, and the links represent idiotypic interactions. A different updating scheme called asynchronous updating was proposed, and it yielded similar results with synchronous updating but with an extended range of influx probability values that does not result to chaotic dynamics. The choice of influx probability remains to affect the system significantly. Memory was observed in the system, especially when the influx probability is low, although no memory cells were incorporated in the network. Time evolution shows the system losing some of its memory, and for systems with high influx probabilities, most of the information is forgotten after some time. Systems with low influx probability continue to hold residual information even after a long time.