Uniting probabilistic methods for optimization

We integrate the following probabilistic search methods: simulated annealing, [a probabilistic version oil tabu search, and [a radically-modified version ofl genetic algorithms. In addition, we use a rejectionless version of simulated annealing, alone or in combination with the other two methods, while in contrast to previous work neither explicitly nor implicitly altering the cooling schedule. We note the implication of this for speed-up when searches are carried out independently on parallel processors.