I present MOSES (meta-optimizing semantic evolutionary search), a new probabilistic modeling (estimation of distribution) approach to program evolution. Distributions are not esti...
Abstract—Dynamic Demes is a new method for the parallelisation of evolutionary algorithms. It was derived as a combination of two other parallelisation algorithms: the master-sla...
— One of the main tasks software testing involves is the generation of the test inputs to be used during the test. Due to its expensive cost, the automation of this task has beco...
Given any measure-preserving dynamical system (Y, A , , T) and g Lp() , we study convergence of the sequence 1 n n k=1 g TSk , n 1 where Sk is a dynamic Zr -valued random w...
—Much of the computational complexity in employing evolutionary algorithms as optimization tool is due to the fitness function evaluation that may either not exist or be computat...
Mohsen Davarynejad, Mohammad R. Akbarzadeh-Totonch...