A Mehrotra-type predictor–corrector infeasible-interior-point method with a new one-norm neighborhood for symmetric optimization
作者：
Highlights：
• We present a Mehrotra-type predictor–corrector infeasible-IPM with a second order corrector step for symmetric optimization.
• We prove the important connection ‖x∘y‖1≤3‖x‖F‖y‖F.
• The proposed algorithm will restrict the iterates to a new one-norm neighborhood.
• We obtain the low theoretical complexity bound for the proposed algorithm.
摘要
•We present a Mehrotra-type predictor–corrector infeasible-IPM with a second order corrector step for symmetric optimization.•We prove the important connection ‖x∘y‖1≤3‖x‖F‖y‖F.•The proposed algorithm will restrict the iterates to a new one-norm neighborhood.•We obtain the low theoretical complexity bound for the proposed algorithm.
论文关键词：Symmetric cones,Euclidean Jordan algebra,One-norm,Infeasible-interior-point method,Complexity bound
论文评审过程：Received 27 November 2012, Revised 3 March 2014, Available online 22 January 2015.
论文官网地址：https://doi.org/10.1016/j.cam.2015.01.027
