Path: utzoo!attcan!uunet!know!zaphod.mps.ohio-state.edu!ub!ubvmsd.cc.buffalo.edu!v091nm4y
From: v091nm4y@ubvmsd.cc.buffalo.edu (Samir Haque)
Newsgroups: comp.ai.neural-nets
Subject: Postsynaptic models
Message-ID: <41918@eerie.acsu.Buffalo.EDU>
Date: 22 Oct 90 03:41:07 GMT
Sender: news@acsu.Buffalo.EDU
Reply-To: v091nm4y@ubvmsd.cc.buffalo.edu
Organization: University at Buffalo
Lines: 17
Nntp-Posting-Host: ubvmsd.cc.buffalo.edu
Keywords:More models?
News-Software: VAX/VMS VNEWS 1.3-4.3



      I have been working on Daniel Alkon's idea of a postsynaptic neural 
      network, which is called DYSTAL. DYSTAL is a postsynaptic net because 
      the learning rule does NOT depend on the firing history of the 
      postsynaptic neuron. This make each neuron somewhat more complex, but  
      the overall complexity is on the o(N) and not o(N^2) or even higher,
      making it one of the most flexible nets available. (N=# of neurons)

      - Are there any other postsynaptic models?
       References or names would be appreciated.

                                             TIA,
                                                 Samir.
========================================================================
"In times of crisis, it is of utmost importantance not to lose one's head"
                                                   -M.Antoinette