A Scaled Central Path for Linear Optimization
The central path is the most important in the design of interior-point algorithm for linear optimization. By an equivalence reformulation for the classical Newton direction, we give a new scaled central path, from which a new search direction is obtained. We derive the complexity bound for the full-step interior point algorithm based on this searching direction and the resulting complexity bound is the best-known for linear optimization.
Helen Zhang, Gang Shen and David Jin
L. P. Zhang and Y. H. Xu, "A Scaled Central Path for Linear Optimization", Advanced Materials Research, Vols. 204-210, pp. 683-686, 2011