Question: NP Complete (Circuits)

timbrigham_at_hotmail.com
Date: 02/23/05 

Date: 23 Feb 2005 09:17:03 -0800

I have developed a program which generates a large random logic
operator (actually a neural network which generates this operator).
This is a feedback driven system; there is no external input to the
system apart from the initial data insertion. The entire system is
cyclic; the output of any given neuron (including those which display
the results) is fed back into the system.

My question is this: With having an exact diagram of the operator and
the output states listed, is determining the state of the system which
creates that output an NP Complete class problem?

Relevant Pages

  • Re: Neural network without training
    ... I want to implement a neural network in which, each neuron ... plane; one neigbour in the higher plane, ... Each connection between a layer above and below ensures ...
    (comp.soft-sys.matlab)
  • matlab webserver and neural network
    ... I'm using matlab for trainning my neural network, which display the ... results to the web browser through matlab webserver. ...
    (comp.soft-sys.matlab)
  • Re: Neural network without training
    ... I want to implement a neural network in which, each neuron ... neighbour in the lower plane. ... Each connection between a layer above and below ensures ...
    (comp.soft-sys.matlab)
  • Re: Error in code when changing number of outputs.
    ... I created a Neural Network two predict the next value of a time series. ... My code is as follows (note that when I use 1 neuron everything is ok): ... % Define neural network input and target ... net = newlin(range, nNeurons, delays, learningRate); ...
    (comp.soft-sys.matlab)
  • Re: The guy who stopped Go Programming Development ...
    ... firings at its input exceeds a treshold, the neuron fires a digital pulse. ... That is the core difference between a "number crunching" system and a neural network, ...
    (rec.games.go)