Adaptive cubic regularisation methods for unconstrained optimization. Part II: worst-case function- and derivative-evaluation complexity2010-01-100 PRE AI DOI:10.1007/s10107-009-0337-y原文链接原文求助分享收藏摘要 En