Average convergence rate of evolutionary algorithms in continuous optimization

The average convergence rate (ACR) measures how fast the approximation error of an evolutionary algorithm converges to zero per generation. It is defined as geometric reduction over consecutive generations. This paper makes a theoretical analysis ACR in continuous optimization. obtained results are summarized follows. According limit property, classified into two categories: (1) linear whose inferior value larger than positive and (2) sublinear zero. Then, it proven that for programming using landscape-adaptive mutation, but landscape-invariant or mutation. relationship between decision space dimension also polynomial reciprocal function any generation, exponential less long period. easy problems such functions, (1+1) adaptive random univariate search polynomial. But hard functions deceptive function, both exponential.