On Sun, Dec 18, 2005 at 02:45:16PM -0500, Pei Wang wrote: > I have no interest in defining neural network. What I want to do is to > formalize a concrete notion of "neural network", which is general > enough to cover many models, and also specific enough so its strength > and weakness can be analyzed in some detail. Eventually I want to
I don't think this can be done. Systems simple enough to analyzable (e.g. single-layer perceptrons) are not powerful. Powerful systems are not analyzable. Frameworks can be spanned wide enough to include systems operating between our ears, but unfortunately lacking knowledge of constraints about said systems operating between our ears the framework is a vast overkill, and certainly not analyzable. > compare it with my own work, so I don't have to do the comparison with > each individual models. > > I understand that the model I gave in the memo looks too narrow to > you. Can you, or anyone else, give me an alternative? What Ben Sure. A spiking system. Average connectivity 10^3. Some 10^3..10^5 automaton classes (cell types). Morphogenetic code for each automaton class which is comparable in power to what neurons and glia do. ~1 ms iteration step. Full physical model of the segment (10^6 segments/automaton). Some 10^10 automatons in toto/system. Spike propagation during processing included origin state, intermediate state and target state (not that much state, less than 16 bit). This is a pretty big envelope, but I'm reasonably certain the system we're looking for is in there. I'm sure the set of constraints will be narrowed further in the process. > proposed is still too general for my purpose. Of course, no such model > can be perfect, but we still need them to make our discussions > meaningful. > > What I don't like in the previous discussion about neural network is: > it is hard to pin down the theoretical assumptions and commitments in > this research paradigm, except the most obvious ones. If everything is > allowed in neural networks, then the concept means nothing. Of course not everything is allowed. But your tool set won't be of much use here. -- Eugen* Leitl <a href="http://leitl.org";>leitl</a> http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE ------- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/[EMAIL PROTECTED]
signature.asc
Description: Digital signature
