An overview of the neuron ring model
The Neuron Ring model employs an avalanche structure with two important distinctions at the neuron level. Each neuron has two memory latches; one traps maximum neuronal activation during pattern presentation, and the other records the time of latch content change. The latches filter short term memory. In the process, they preserve length 1 snapshots of activation theory history. The model finds utility in pattern classification. Its synaptic weights are first conditioned with sample spectra. The model then receives a test or unknown signal. The objective is to identify the sample closest to the test signal. Class decision follows complete presentation of the test data. The decision maker relies exclusively on the latch contents. Presented here is an overview of the Neuron Ring at the seminar level.