BEGIN:VCALENDAR VERSION:2.0 PRODID:-//Memento EPFL// BEGIN:VEVENT SUMMARY:Distributed Optimization via the Alternating Direction Method of M ultipliers DTSTART:20130712T101500 DTEND:20130712T111500 DTSTAMP:20260406T172159Z UID:da83ea3b981e2e2bb9bf5700661ebba5eb8763a5a504dee5e2525755 CATEGORIES:Conferences - Seminars DESCRIPTION:Prof. Stephen P. Boyd\, Stanford University\nBio: Stephen P. B oyd is the Samsung Professor of Engineering\, and Professor of Electrical Engineering in the Information Systems Laboratory at Stanford University. He has courtesy appointments in the Department of Management Science and E ngineering and the Department of Computer Science\, and is member of the I nstitute for Computational and Mathematical Engineering. His current resea rch focus is on convex optimization applications in control\, signal proce ssing\, and circuit design.\nProfessor Boyd received an AB degree in Mathe matics\, summa cum laude\, from Harvard University in 1980\, and a PhD in EECS from U. C. Berkeley in 1985. In 1985 he joined the faculty of Stanfor d's Electrical Engineering Department. He has held visiting Professor posi tions at Katholieke University (Leuven)\, McGill University (Montreal)\, E cole Polytechnique Federale (Lausanne)\, Tsinghua University (Beijing)\, U niversite Paul Sabatier (Toulouse)\, Royal Institute of Technology (Stockh olm)\, Kyoto University\, Harbin Institute of Technology\, NYU\, and MIT. He holds an honorary doctorate from Royal Institute of Technology (KTH)\, Stockholm.\nProfessor Boyd is the author of many research articles and thr ee books: Convex Optimization (with Lieven Vandenberghe\, 2004)\, Linear M atrix Inequalities in System and Control Theory (with L. El Ghaoui\, E. Fe ron\, and V. Balakrishnan\, 1994)\, and Linear Controller Design: Limits o f Performance (with Craig Barratt\, 1991). His group has produced several open source tools\, including CVX (with Michael Grant)\, a widely used par ser-solver for convex optimization.\nProblems in areas such as machine lea rning and dynamic optimization on a large network lead to extremely large convex optimization problems\, with problem data stored in a decentralized way\, and processing elements distributed across a network. We argue that the alternating direction method of multipliers is well suited to such pr oblems.  The method was developed in the 1970s\, with roots in the 1950s\ , and is equivalent or closely related to many other algorithms\, such as dual decomposition\, the method of multipliers\, Douglas-Rachford splittin g\, Spingarn's method of partial inverses\, Dykstra's alternating projecti ons\, Bregman iterative algorithms for l_1 problems\, proximal methods\, a nd others.  After briefly surveying the theory and history of the algorit hm\, we discuss applications to statistical and machine learning problems such as the lasso and support vector machines\, and to dynamic energy mana gement  problems arising in the smart grid. LOCATION:CO 3 http://plan.epfl.ch/?room=CO%203 STATUS:CONFIRMED END:VEVENT END:VCALENDAR