GenerateCrossesfromElites: Generate crosses from elites
In STPGA: Selection of Training Populations by Genetic Algorithm

Given a list of elite training sets, list of candidates the function makes npop new solutions by using crossover and mutation operators.

1 2	GenerateCrossesfromElites(Elites, Candidates, npop, mutprob, mc.cores = 1, mutintensity = 1, memoryfortabu = NULL)

`Elites`	a list of elite training sets
`Candidates`	a vector of identifiers of the individuals in the candidate set.
`npop`	number of training sets to generate.
`mutprob`	point mutation probability for each individual generated. Only one mutation per solution is allowed.
`mc.cores`	number of cores to use.
`mutintensity`	mean of the poisson variable that is used to decide the number of mutations for each cross.
`memoryfortabu`	tabu memory