Building evolution friendliness into cellular automaton dynamics: the cytomatrix neuron model

The cytomatrix neuron is a softened cellular automaton, roughly motivated by interactions that could occur in a molecular or cellular complex. Input signals are combined in space and time by subcells that exert graded influences on each other. Output is triggered if a readout element is located in a suitably activated subcell. Multiple parameters are open to evolution. Extensive experimentation with the model shows that the dynamics can be molded to produce different structures of generalization. Dimensionality can be increased by increasing the number of dynamical parameters open to variation and selection. Learning algorithms that vary the greatest number of parameters were found to have a greater variability in the structures of generalization and to yield higher performance values and learning rates. Here we focus on n-bit exclusive-OR tasks that are known to be hard due to their linear inseparability. The system successfully learned 2 bit and 4 bit X-OR functions. The higher dimensional algorithms exhibited a relatively good performance on the 8 bit X-OR function