Theoretically, the asymptotic stability of training pattern and a lower bound of attraction basin are discussed. 在理论上,讨论了训练模式的渐近稳定性和吸引域的下界。
A new lower bound of the problem is given for the further evaluation of the heuristic numerically. The asymptotic optimality of the lower bound is proved and its worst case ratio is the number of machines m, and this bound is tight. 为了进一步从数值上对提出的启发式进行评价,提出了一个新的下界,证明了该下界的渐近最优性,并指出其最坏情况比为机器数m,且为紧界。