Department of Computer Science | Institute of Theoretical Computer Science | CADMO

Theory of Combinatorial Algorithms

Prof. Emo Welzl and Prof. Bernd Gärtner

Mittagsseminar (in cooperation with A. Steger, D. Steurer and B. Sudakov)

Mittagsseminar Talk Information

Date and Time: Tuesday, April 03, 2012, 12:15 pm

Duration: 30 minutes

Location: OAT S15/S16/S17

Speaker: Sebastian Stich

The Heavy Ball Method

The classical gradient method is one of the main pillars of the successful story of modern convex optimization. Complexity analysis has revealed that the convergence rate of the gradient method is far from the theoretical lower bound. Although optimal methods have already been proposed in the 80s, they became popular not before the last decade.

We analyze the convergence rate of the gradient method on smooth strongly convex functions and present the lower complexity bound for this class. We introduce Polyak's Heavy Ball Method. This is one of the simplest multistep schemes which does not only have an easy physical explanation, but also achieves optimal convergence rate.

Upcoming talks     |     All previous talks     |     Talks by speaker     |     Upcoming talks in iCal format (beta version!)

Previous talks by year:   2024  2023  2022  2021  2020  2019  2018  2017  2016  2015  2014  2013  2012  2011  2010  2009  2008  2007  2006  2005  2004  2003  2002  2001  2000  1999  1998  1997  1996  

Information for students and suggested topics for student talks

Automatic MiSe System Software Version 1.4803M   |   admin login