Acousto-conformational phase transitions in the cytoskeleton, adaptive resonance networks with nonlinear synapses and trainable intraneuronal pattern recognition

摘要

A possible role for phonon excitations of the cytoskeleton in intraneuronal pattern recognition and learning is discussed. Biophysical support is presented for molecular implementation of adaptive resonant theory neural network principles. Relations between pattern recognition by neural network and symmetry breaking second order phase transitions are discussed.