The Evolution of Learning Algorithms for Artificial Neural Networks Published 1992 in Complex Systems by Jonathan Baxter Michael Tauraso
Jan 14, 2016
The Evolution of Learning Algorithms for Artificial
Neural Networks
Published 1992 in Complex Systems by Jonathan Baxter
Michael Tauraso
Genetic Algorithm on NNsStart with a population of neural networks.Find the fitness of each for a particular taskWeed out the low-fitness onesBreed the high-fitness ones to make a new
population.
Repeat.
Local Binary Neural Networks(LBNNs)
All weights, inputs, and outputs are binary.Learning rule is a localized boolean function of
two variables.This vastly simplifies everything.LBNNs are easy to encode into binary strings.LBNNs are easy to write into genetic
algorithms
An LBNN
Rules for LBNNs Weights are +1, -1, or 0 Nodes: ai(t+1) =sign( ∑ aj(t)wji(t) )
Weights: wij(t+1) = f(ai(t), aj(t))
Weights are classified as fixed or learnable. 0 weights are fixed.
Training RulesBoolean functions of two variables16 possible varietiesAnalog of Hebb’s rule given by:
f(ai(t),aj(t)) = ai(t) aj(t)
Training GoalLearn the 4 boolean functions of one variable Identity, Inverse, Always 1, Always 0Who wants to learn the boolean functions of one
variable anyway?
Fitness DeterminationStart with an LBNN from the sample
populationClamp the output node to train for a
particular boolean function.Fitness is how well the network performs at
calculating that boolean function after training.
A Successful LBNN
FindingsHebb’s rule is the most efficient learning
rule.LBNNs can be thought of as state
machines
LBNNs as State MachinesBoolean functions are encoded as transitions
between fixed points in the NNOther transitions seek to push the network
toward the appropriate fixed point.
State Machine for an LBNN
Questions
?