Prof. Emo Welzl and Prof. Bernd Gärtner
|Mittagsseminar Talk Information|
Date and Time: Tuesday, September 23, 2003, 12:15 pm
Duration: This information is not available in the database
Location: This information is not available in the database
Speaker: Vinayak Pandit (IBM India)
The local search heuristics are very popular among practitioners in the area of operations research, pattern recognitions, and related areas. Some of the famous problems for which local search is a method of choice include travelling salesman problem, k-median problem, and k-means clustering. Unfortunately, most local search heuristics are not easy to analyse, and prove properties about optimality, or worst case analysis. In this talk, we present an analysis of the local search heuristics for the k-median problem.
Significantly, this analysis yields the best known approximation ratio for k-median problem. In my talk, I will consider a simple local search heuristic for the k-median problem, and prove that its local optimal is at most 5 times more than the global optimal. Our analysis also works for facility location problems.
Automatic MiSe System Software Version 1.4803M | admin login