Overview | HNet | Performance Aspects | Mathematics | Biology


Commutativity is a very powerful aspect of holographic processing.  Learning capacity increases in linear proportion to the number of cortical memory elements across all connected neuron cells.  This has important implications for the neo-cortical model.

Dynamic Signal Routing

A subset of HNeT cell types are designed specifically for routing of axonal signals among cell assemblies.  These perform functions analogous to the association and thalamocortical projection fibers of the midbrain.

Dynamic Cell Restructuring

The structure of HNeT cells are dynamic and may be modified while execution proceeds in real time.  These aspects concern physical cell processes and internal structures such as axo-dendritic connections, number of axons / dendrites / cortical memory elements, and other aspects such as learning rate and memory decay.

Hyperincursive Learning

The basis of self-referential or cognitive systems, this forms the basis for the HNeT unsupervised learning model.

Signal Preprocessing

The HNeT system is supplied with a broad range of standard input / output signal conversion methods including sigmoid, Fourier, Gabor, wavelet, quantization, windowing, etc.