Two simple NeuroCognitive associative memory models

Human memory is associative and emerges from the behaviour of neurons. Two models, based on commonly used biological neural models are presented. The first model uses static synapses to approximate timing behaviour for a Stroop task with congruent conditions responding faster than incongruent conditions. The second model uses plastic synapses to learn a semantic net; it then duplicates the behaviour of a question answering task. This behaviour not only answers correctly, its times are similar to that of human subjects. These models are flawed in many ways, for instance, they use hundreds of neurons instead of the billions of neurons in the brain. They are thus not proposed as anything near a complete final model, but instead as early steps toward the development of more sophisticated neurocognitive associative memory models.

(C) Copyright 2018 retained by the authors.