Bruce has had a website up continuously since 1994 (see History) and has published a blog since 2003. He calls Godzone home but has worked mostly in the UK, practicing medicine with an interest in health informatics (writing 3 books). Following a reflective “gap year” in Wellington, the capital of New Zealand and Heart of the Edge of the World, he elected to relinquish clinical work and now works in a public health role. He enjoys photography, travel, using Macs, tinkering with various gadgets, co-founded the Geotag Icon Project and co-designed the Open Share Icon. This site is partly hobby, partly memory aid, and partly because some stuff is worth sharing.
It sounded cool and the domain wasn’t taken… Neural networks owe their origins as a topic of study to work at Stanford University in the 1950s. A neural network is a crude type of artificial intelligence in which multiple connections link to the processing units (nodes or artificial neurons) in parallel. The number and arrangement of these inter-neural connections weights the output of a decision made at a unit/ neurone so that it can vary from being simply true or false (0 or 1). The biological inspiration for neural networks is readily apparent. Neural networks are strictly-speaking non-digital because they are not limited to manipulating zeros and ones. They can, however, be simulated on computers—which are of course digital. These digital simulations are properly referred to as artificial neural networks or ANNs. Non-digital, biological neural networks (bioneural nets) are considerably more complicated than ANNs. Existing examples (e.g. the neurons and dendritic connections of the brain) can be studied to supply inspiration, but thus far cannot be duplicated.