This paper presents a new primal-dual interior-point algorithm with reduced potential function for a class of convex programming, based on the ideas of that method for solving linear programming. The new algorithm chooses the classical Newton direction as iteration direction and its iteration stepsize is determined by potential function. As the search directions Δx and Δs aren't orthogonal any more, the complexity analysis of this method is different from that of linear programming, correspondingly. Under a scaled Lipschitz condition, the algorithm is proved to possess O(nL) iteration-complexity bounds © 2008 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Zhao, Y., & Zhang, M. (2008). A wide neighborhood primal-dual interior-point algorithm for a class of convex programming. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5227 LNAI, pp. 734–741). https://doi.org/10.1007/978-3-540-85984-0_88
Mendeley helps you to discover research relevant for your work.