Study on Randomness in Genetic Algorithm Evolution
随机性对遗传算法演化机制的影响

Tries to find to what extent GAs relies on random mechanism. This article brings forward 3 algorithms with ascending randomness and compares their searching abilities with simple GA with Elitist algorithms, whose random degree is the highest. Four fitness functions are proposed as test beds for these algorithms. Results show that that the search ability of TIGA algorithm, whose random degree is neither the highest nor the lowest, is the best among the tested sister algorithms.