CAs recognise quickly
- CAs have a base level of activation; it's background noise.
- When some neurons in the CA are activated externally, internal
activation increases and many of the neurons in the CA are activated.
- When a significant percentage of the neurons are activated,
the CA "recognises" the external input as an instance of the
CAs category.
- This follows the "snoopy" curve.
- This is derived from Kaplan's theoretical work. My practical
work shows peak activation as fast as cycle 3, (30 ms) and
typically around cycle 10.
- I suspect these numbers are too quick. The model doesn't acurately
represent our own classification.
- These models can classify things really quickly.