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, March 31, 2015, 12:15 pm

Duration: 30 minutes

Location: OAT S15/S16/S17

Speaker: May Szedlák

Combinatorial Redundancy Detection Algorithm

The problem of detecting (and removing) redundant constraints is fundamental in optimization. We focus on the case where we are given a set H of n halfspaces in the d-dimensional real space. The feasible solution set is given by the intersection of all halfspaces in H and a halfspace is called redundant if its removal does not change the feasible solution set. The currently fastest known algorithm to detect all redundancies is the one by Clarkson. This method solves n linear programs, each of them on at most s variables, where s is the number of nonredundant variables. In this talk we study the combinatorial aspect of redundancy detection. How and how fast can we detect all redundant halfspaces? Instead of the linear system we only consider the finitely many signed dictionaries, i.e., matrices that can be thought of as an enriched version of an intersection point of d halfspaces of H. We show that given only this combinatorial information, there is an output sensitive algorithm to detect all redundancies. Although our running time is worse than Clarkson's, in the case where all constraints are in general position we essentially match its running time.

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