The challenge of taking many variables into account in optimization problems
may be overcome under the hypothesis of low effective dimensionality. Then, the
search of solutions can be reduced to the random embedding of a low dimensional
space into the original one, resulting in a more