The Use of an Analytic Quotient Operator in Genetic Programming

We propose replacing the division operator used in genetic programming with an analytic quotient (AQ) operator. We demonstrate that this AQ operator systematically yields lower mean squared errors over a range of regression tasks, due principally to removing the discontinuities or singularities that can often result from using either protected or unprotected division. Further, the AQ operator is differentiable. We also show that the new AQ operator stabilizes the variance of the intermediate quantities in the tree.