pitts v0.1.0

McCulloch Pitts Neurons

This library implements simple threshold logic in the style of McCulloch-Pitts Neurons. The function of the neurons and the examples are dervied from the book Computation: Finite and Infinite Machines by Marvin Minsky.

Example

A simple latch. This network will continue to output a 1 after (t+1) the first 1 appears in the input. An additional reset input will return the network to a quiescent state.

n = new Pitts.Network()

input = n.addInput [0,0,1,0, 0]
reset = n.addInput [0,0,0,0,0,0,1]
node_a = n.addNeuron 1
output = n.addOutput()

n.connect(input, node_a)
n.inhibit(reset, node_a)

n.connect(node_a, node_a)
n.connect(node_a, output)

More examples found under examples/

License: BSD