> > > > > Yeap, there's well developed theories about how an autoassociate > > network like CA3 could support multiple, uncorrelated attractor > > maps and sustain activity once one of them was activated. The > > big debate is about how they are formed. > > The standard way attractors are formed in formal ANN theory is via variants > of Hebbian learning. But pure unsupervised Hebbian learning has never > worked very well in simulations. In the CS theory of reinforcement > learning, a lot of tricks have been used to make Hebbian learning work > better (temporal difference learning, for example), but none of these work > that awesomely. >
Using artificial rules, such as hardball winner-take-all and synaptic weight normalization, it's doable to get ANN's to do this. But in an autoassociative network with realistic biophysical properties, controlling activity to prevent runaway synaptic modification is a very large problem. My own grad advisor, Mike Hasselmo has worked on this very problem using pharmacological modulation to suppress synaptic transmission during learning. The fact that epilepsy usually starts within the hippocampus (with its sheets of 100k neurons, all interconnected with excitatory connections) indicates that this is a real problem for the brain as well as models of it. I imagine the hippocampus is really pushing the evolutionary envelope in terms of being prone to epilepsy. Demand for more memory is probably fighting directly against epileptic tendencies in terms of evolutionary fitness. Another problem is feeding that dense sheet of nerves (which is why the hippocampus is one of the first things to suffer damage during anoxia). It's a very specialized area that's pushing the limits of the body's ability to feed it and keep it from siezing up. > > Do you think the spike-time data contains enough information that it's not > necessary to look for patterns in the raw data? They keep the real data too, but it's *huge* (100+ channels of 70khz data, realtime). The raw data is basically an average of the neural activity of the nearby cells. Spikes from neurons within a small radius of the electrode tip stand out and have a certain characteristic shape/amplitude, which is used to identify said cell. Apart from identifying spikes, I'm not sure you'd get much out of the raw data(assuming you are also collecting EEG data realtime at 10khz or so from one electrode in the nearby region). However, nowadays people are starting to worry about complex spikes too (bursts of spikes). Assigning these spikes to their source neuron is much harder because spikes after the first one in a burst are reduced in amplitude. So you need specialized clustering algorithms that are aware of bursts and what they do to a spike amplitude. You need to go back to the raw data to identify such bursts every time you change your detection algorithm. -Brad ------- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/[EMAIL PROTECTED]
