Source: Comp.ai.neural-Nets
"Note, in all the mountain climbing methods currently discussed, kangaroo is most hoped to reach the top of the top of its starting point. But not to ensure that the top of the mountain is Everest, or a very high peak, all methods of use are trying to find the actual global global The optimal value.
In the SA (an annealing algorithm), the kangaroo was drunk, and randomly jumped for a long time, but it gradually woke up and jumped toward the peak.
In GA (Genetic Algorithm), there are many kangaroos, they landed in the Himalayas (if the pilots are not lost). These kangaroo don't know that they are imagined looking for Zhu Mount, but every few years, you will shoot some kangaroo highly low places, and hope to survive those who are multi-origin, and have children. " .
