What’s new?
On 19/20 October 2006 I successfully defended my PhD thesis at the University of Fribourg. The final, rather clunky (IMO) title, was Physics-based approaches to the dynamics of learning, brains and socioeconomic systems. I'm now working for a while with the group of Andrzej Nowak at the University of Warsaw, helping with some interesting economics experiments.
This site is eternally, always, under construction. >;}~
It describes results in neural systems research by me (Joseph Wakeling). At the time of writing I am pursuing a PhD in Theoretical Physics at the University of Fribourg, Switzerland, after graduating from Imperial College, London, with an MSci in Mathematics. As well as my own research, you can find here links and information on other related papers and subjects that I feel are interesting or relevant, including AI, cognitive science and related physics research.
What do physicists have to contribute to the understanding of neural systems? The aim is to gain an understanding of the fundamental mechanics that are at work. Neuroscientists have identified the ‘macro’ regions of the (human) brain responsible for many aspects of behaviour and/or experience — vision, hearing, touch, smell etc. — but macro regions of the brain differ from organism to organism. What is it that links the brain of a human with that of an octopus, that of a lobster, that of a lizard? In particular, what is it about the basic interactions of individual neurons that produces the overall behavioural patterns that we see?
The last 15 years have seen the creation and rapid expansion of a new idea in physics known as Self-Organized Criticality (SOC), the study of non-equilibrium systems displaying behaviour that is finely balanced between order and randomness. This has turned out to be a powerful tool for gaining an understanding of many natural systems: avalanches, earthquakes, solar flares, pulsars, the formation of river networks and the spread of forest fires are all now much better understood in the light of SOC theory. A general characteristic of all SOC systems is that, while they display complex global (macro) behaviour, this is the result of simple interactions at the micro level. Thus, SOC seems like a good candidate for explaining the underlying mechanics of neural systems.
My own small contribution to this field are some papers building on work by Per Bak, the founding father of SOC, and the Argentinian neuroscientist Dante Chialvo. Together they developed an elegant little neural network model which I rather cheekily dubbed the ‘minibrain’ (Per later told me that he always called it the ‘Dante brain’, and Dante called it ‘Per learning’ — you can draw your own conclusions!-). Most of the work here centres around investigating different aspects of this model.
So that’s the introduction, where’s the meat? First and foremost there are the papers, links to papers by me and papers by others that I find interesting. You’ll find a complete history of minibrain-related stuff here. There is also source code for computer experiments, and a page of links to lots of different fun neural systems related stuff. Take it all with a pinch of salt, have fun, and let me know what you think!