Searching neural network structures with L systems and genetic algorithms

We present a new method for using genetic algorithms and L systems to grow up efficient neural network structures. Our L rules operate directly on 2-d imensional cell matrix, L, rules are produced automatically by genetic algo rithm and they have "age" that controls the number of firing times, i.e., t imes we can apply each rule. We have modified the conventional neural netwo rk model so that it is easy to present the knowledge by birth (axon weights ) and the learning by experience (dendrite weights). A connection is shown to exist between the axon weights and learning parameters used e.g., in bac k propagation. This system enables us to find special structures that are v ery fast for both to train and to operate comparing to conventional, layere d methods.