Main»Approx Seminar

Approximation Algorithms Seminar

Moses Charikar

This is a weekly seminar focused on approximation algorithms. Initial meetings will be devoted to building background material and in later meetings, we will read recent papers in the area.

Time: Wednesday, 10:30 am Location: CS 302

Mailing list for announcements:

  • Subscribe to approx-seminar.
  • [Feb 8] Introduction to approximation algorithms, and some intriguing open problems.
    • Vertex cover
    • Max acyclic subgraph
    • Scheduling unit time jobs
    • TSP
  • [Feb 15] LP relaxations, duality and lagrangean relaxation
  • [Feb 22] Cut problems, metric methods, divide and conquer
    • I talked about metric methods, specifically about multicut and balanced cut.
    • For background, see the notes from lecture 11 in David Williamson's notes on Approximation algorithms.
  • [Mar 1] spreading metrics, divide and conquer
    • I will talk about uses of cut procedures in divide and conquer algorithms, specifically in minimum linear arrangement and feedback arc set.
    • For background, see the notes above.
  • [Mar 8] Introduction to SDP: max cut and graph coloring
  • [Mar 15] Eddie
    • New Approximation Guarantees for Chromatic Number, to appear in STOC 2006.
  • [Mar 29] primal-dual algorithms for connectivity
    • See Sections 7.1-7.4 of draft Approximation Algorithms text
    • Chapter by Goemans and Williamson in Hochbaum's book.
  • [Apr 5] No seminar (Moses out of town)
  • [Apr 12] Mohammad
  • [Apr 19] Approximation schemes for geometric problems
    • See Sanjeev's survey. This survey accompanied a semiplenary talk at ISMP and appeared in Math Programming, 97 (1,2) July 2003.
    • This is a precursor to Wolfgang's presentation next week on a paper on min-weight triangulation (in STOC 2006).

Possible topics for future weeks

  • Primal dual method
    • generalized steiner tree
  • iterative rounding
    • survivable network design
  • local search
    • k-median
    • facility location via dual fitting
  • cost sharing schemes
    • stochastic optimization
  • SDP methods
    • Max Cut
    • Coloring
    • l_2^2 metrics
  • Dynamic programming
    • Knapsack
    • Eucidean TSP
    • Minimum triangulation in the plane
  • Tree approximations
    • distance approximation by HSTs
    • distance approximation by spanning trees
    • capacity approximation by tree decomposition

Suggestions for presentations