When using an evolutionary algorithm on an unknown problem, properties
like the number of global/local optima must be guessed for properly picking
an algorithm and its parameters. It is the aim of current paper to put forward
an EA-based method for real-valued optimization to provide an estimate on the
number of optima a function exhibits, or at least of the ones that are in reach for a
certain algorithm configuration, at low cost.We compare against direct clustering
methods applied to different stages of evolved populations; interestingly, there is
a turning point (in evaluations) after which our method is clearly better, although
for very low budgets, the clustering methods have advantages. Consequently, it is
argued in favor of further hybridizations.