1 / 6

Second Year Viva

Second Year Viva. Chris Unsworth. What I have done. A new description for SMN. Why? To fit more intuitive into an AC5 style arc-consistency algorithm. As a blueprint for future constraint discriptions. A new complexity argument to fit the new description. Proofs

orien
Télécharger la présentation

Second Year Viva

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. Second Year Viva Chris Unsworth

  2. What I have done • A new description for SMN. • Why? • To fit more intuitive into an AC5 style arc-consistency algorithm. • As a blueprint for future constraint discriptions. • A new complexity argument to fit the new description. • Proofs • soundness and completeness • Equivalence to the EGS algorithm • Failure free enumeration of all stable matchings

  3. What I have done • Empirical study of constraint solutions to the stable marriage problem. • Solutions included: • Conflict matrix encoding • Boolean encoding • N-Valued encoding • 4-Valued encoding • SM2 constraint • SMN constraint • EGS algorithm • Instance sizes and sample sizes • 10 – 100 sample size 1000 • 100 – 1000 sample size 1000 • 1000 – 8000 sample size 20 • SMN is comparable to EGS • when n >= 1000, EGS at most 4 times faster than EGS

  4. What I have done • Comparing the average number of solutions lead the conjecture that: “after arc-consistency has been enforced over a constraint model which uses the SMN constraint, the first stable matching can be found in n time and each subsequent solution can also be found in n time”

  5. What I am doing • Starting 17th October 2005 • Check implementation of all CP SM models, including: • FT, Bool, DG1, DG2, SM2 and SMN • Get a reasonable implementation of the EGS algorithm in Java • Design, run and tabulate an empirical study of the above models • Starting November 2005 • Check implementation of SMTI2 • Implement SMTIN • Recreate ECAI 02 experiments and include the new models • Increase the instance sizes to explore the limits of the new solutions • Write up the work with a view to submitting to ECAI 06 (deadline for summaries 8th February 2006)

  6. What I am going to do • Provisional thesis statement • “A Specialised constraint solution for a stable matching problem can solve the problem in comparable time to an algorithmic solution, and it can significantly outperform a more traditional toolbox constraint solution whilst maintaining its versatility ”

More Related