Models of competitive learning: complex dynamics, intermittent conversions, and oscillatory coarsening

We present two models of competitive learning, which are respectively interfacial (i.e., where the interiors of domains are unaffected by the rules of the model) and cooperative (i.e., where the bulk as well as the interface of an individual domain is governed by the rules of the model) learning. This learning is outcome related, so that spatially and temporally local environments influence the conversion of a given site between one of two different types. We focus here on the behavior of the models at coexistence, which yields critical behavior and the existence of a phase involving a type of coarsening which is oscillatory in nature. In the discussion we speculate on, among other issues, the likely behavior of the models away from coexistence.