1 / 22

RandomEdge can be mildly exponential on abstract cubes

RandomEdge can be mildly exponential on abstract cubes. Tibor Szab ó ETH Z ü rich. Jiri Matousek Charles University Prague. Linear Programming.

amato
Download Presentation

RandomEdge can be mildly exponential on abstract cubes

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. RandomEdge can be mildly exponential on abstract cubes Tibor Szabó ETH Zürich Jiri MatousekCharles University Prague

  2. Linear Programming • Given a convex polyhedronP in Rn with at mostm facets and a linear objective function c, one would like to determine the minimum value ofconP. • The minimum is taken at a vertex of P. • The simplex algorithm moves from vertex to vertex along an edge each time decreasing the objective function value. • The way to select the next vertex is the pivot rule

  3. RandomEdge • RandomEdge is the simplex algorithm which selects an improving edge uniformly at random. • Its running time • on the d-dimensional simplex is Liebling • on d-dimensional polytopes with d+2 facets is Gärtner et al. (2001) • on the n-dimensional Klee-Minty cube is Williamson Hoke (1988) Gärtner, Henk, Ziegler (1995) Balogh, Pemantle (2004)

  4. Abstract Objective Functions • P is a polytope • f : V(P) → Ris an abstract objective function if a local minimum of any face F is also the unique global minimum of F. Adler and Saigal, 1976. Williamson Hoke, 1988. Kalai, 1988.

  5. RandomFacet on AOF • Kalai (1992): the simplex algorithm RandomFacet finishes in subexponential time on any AOF. (also: Matousek, Sharir and Welzl in a dual setting) • Matousek gave AUSOs on which Kalai’s analysis is essentially tight.

  6. RandomEdge is quadratic on Matousek’s orientations • Williamson Hoke (1988) conjectured that RandomEdge is quadratic on all AOFs.

  7. Acyclic Unique Sink Orientations • Let P be a polytope. An orientation of its graph is called an acyclic unique sink orientation or AUSO if every face has a unique sink (that is a vertex with only incoming edges) and no directed cycle. • AUSOs and AOFs are the same

  8. Killing RandomEdge Theorem. There exists an AUSO of the n-dimensional cube, such that RandomEdge started at a random vertex, with probability at least , makes at least moves before reaching the sink.

  9. Ingredients of the good pasta Ingredients of a slow cube • The flour: • The water: • The eggs: • The mixing: Klee-Minty cube Blowup construction Hypersink reorientation Randomness

  10. Klee-Minty cube reversed KMm-1 KMm KMm-1

  11. Blowup Construction

  12. Hypersink reorientation

  13. A simpler construction Let A be an n-dimensional cube, on which RandomEdge is slow. Let . • Take the blowup of Awith random KMm whose sink is in the same copy of A • Reorient the hypersink by placing a random copy of A.

  14. rand A A simpler construction A A A A

  15. A typical RandomEdge move v • Move in frame: • RandomEdge move in KMm • Stay put in A • Move within a hypervertex: • RandomEdge move in A • Move to a random vertex of KMm on the same level A A A rand A Random walk with reshuffles on KMm RandomEdge on A

  16. Walk with reshuffles on KMm • Start at a random v(0) of KMm • v(i)is chosen as follows: • With probability pi,stepwe make a step of RandomEdge from v(i-1). • With probability pi,reshwe reshuffle the coordinates of v(i-1) to obtain v(i) . • With probability 1-pi,step -pi,resh, v(i) =v(i-1).

  17. Walk with reshuffles on KMm is slow Proposition. Suppose that Then with probability at least The random walk with reshuffles makes at least steps. (αandβare constants)

  18. Reaching the hypersink Either we reach the sink by reaching the sink of a copy of A and then perform RandomEdge on KMm. This takes at least T(n) time. Or we reach the hypersink without entering the sink of any copy of A. That is the random walk with reshuffles reaches the sink of KMm. This takes at least time.

  19. The recursion • RandomEdge arrives to the hypersink at a random vertex. Then it needs T(n) more steps. So passing from dimension n to n+nthe expected running time of RandomEdgedoubles. Iterating n - times gives • In order to guarantee that reshuffles are frequent enough we need a more complicated construction and that is why we are only able to prove a running time of .

  20. Open questions • Obtain any reasonable upper bound on the running time of RandomEdge • Can one modify the construction such that the cube is realizable? (I don’t think so …) • Or at least it satisfies the Holt-Klee condition? • Or at least each three-dimensional subcube satisfies the Holt-Klee condition?

  21. More open questions • The model of unique sink orientations of cubes (possibly with cycles) include LP on an arbitrary polytope. Find a subexponential algorithm.

  22. THE END

More Related