CMD_11

Welcome to the Department of Mathematics

file
file

Daily News

Upcoming Events

  • PIMS Analytic RNA combinatorics meeting
    9:00 AM - 12:00 PM
    April 16, 2014
    Current RNA bioinformatics strongly relies on discrete abstractions and decomposition schemes for the conformation of nucleic acids. This workshop will bring together scientists from bioinformatics, mathematics and computer science in an attempt to bridge the gap between the enumeration and (asymptotic) analysis of models for RNA structure, and the design of RNA algorithms. http://tinyurl.com/parc-2014 contact: mmishna@sfu.ca yann.ponty@lix.polytechnique.fr
  • O.R. Seminar: Wotao Yin (UCLA)
    2:30 PM - 3:30 PM
    April 17, 2014
    Title: Distributed Optimization over Network Abstract: There has been considerable recent interest in solving optimization problems with data stored over a network. For these problems we need techniques that process data locally yet converge rapidly to an (approximate) solution across the entire network. This talk reviews primarily first-order algorithms for large-scale optimization of the distributed or decentralized types. We emphasize on recognizing separable structures in a large set of signal processing and statistical learning problems and demonstrate that, through skillful uses of gradient, proximal, duality, and splitting techniques, massively parallel algorithms can be developed. Numerical results are presented to demonstrate the scalability of the parallel codes on typical Unix clusters and Amazon EC2.
  • SFU Discrete Math Seminar (Brad Jones SFU)
    2:30 PM - 3:30 PM
    April 22, 2014
    Title: Hook length formulae, Butcher series and tree factorials
  • PIMS-CSC Distinguished Speaker Series
    3:30 PM - 4:30 PM
    April 25, 2014
    PIMS-CSC Distinguished Speaker Series Friday, April 25, 2014 IRMACS Theatre 10900 3:30 pm (Pre-talk reception at 3:00 pm) Speaker: L. Mahadevan, Harvard University Title: tba Abstract: tbd
  • SFU Discrete Math Seminar (David Bryant Otago)
    2:30 PM - 3:30 PM
    April 29, 2014
    Title: The Geometry of Hypergraphs The work of Linial, London and Rabinovich on the Geometry of Graphs has had a significant and long-lasting impact on combinatorial optimization and data analysis. They showed how the mathematics of metric embedding could help solve difficult computational problems, adding some powerful techniques to the algorithm designer's toolkit. I will show how much of this theory extends seemlessly to diversities, a simple generalization of metric spaces beyond pairwise comparisons. The concept of a diversity arose from an unlikely blend of phylogenetics, metric geometry and category theory, and has generated a rich theory which we are only just beginning to explore. I'll talk about our results in diversity embeddings and how diversities provide the natural geometry of hypergraphs.
  • Stephen Melczer, M.Sc. Thesis Defence, Mathematics Room: K9509
    11:00 AM - 11:30 AM
    May 5, 2014
    Sr. Supervisor: Marni Mishna Title: Variants of the Kernel Method for Lattice Path Models Abstract: The kernel method has proved to be an extremely versatile tool for exact and asymptotic enumeration. Recent applications in the study of lattice walks have linked combinatorial properties of a model to algebraic conditions on its generating function, demonstrating how to extract additional information from the process. This thesis details two new results. In the first, we apply the iterated kernel method to determine asymptotic information about a family of models in the quarter plane, finding their generating functions explicitly and classifying them as non D-finite. The second considers d-dimensional walks restricted to an octant whose step sets are symmetric over every axis. A generalized version of the orbit sum method allows for a representation of their generating functions as diagonals of multivariate rational functions, proving they are D-finite. In combination with current developments from analytic combinatorics in several variables, this yields dominant asymptotics for all such models.
  • Download .ics