Smoothed Analysis of InteriorPoint Algorithms Termination.pdf

Smoothed Analysis of Interior-Point Algorithms: Termination Daniel A. Spielman ∗ Shang-Hua Teng † Department of Mathematics Department of Computer Science 3 0 Massachusetts Institute of Technology Boston University and 0 Akamai Technologies Inc. 2 n February 1, 2008 a J 1 2 Abstract ] S D We perform a smoothed analysis of the termination phase of an interior-point method. . By combining this analysis with the smoothed analysis of Renegar’s interior-point algorithm s in [DST02], we show that the smoothed complexity of an interior-point algorithm for linear c 3 [ programming is O(m log(m/σ)). In contrast, the best known bound on the worst-case 3 complexity of linear programming is O(m L), where L could be as large as m. We include 1 an introduction to smoothed analysis and a tutorial on proof techniques that have been v 9 useful in smoothed analyses. 1 0 1 1 Introduction 0 3 0 This paper has two objectives: to provide an introduction to smoothed analysis and to present / s a new result—the smoothed analysis of the termination of interior-point algorithms. We begin c : with an intuitive introduction to smoothed analysis (Section 1.1) followed by a more formal v i introduction (Section 1.2). After introducing necessary notation in Section 2, we survey the X complexity of interior-point algori