|
Tuesday 06 January 2009
Login |
Recent and Upcoming SeminarsTo be announced. Page 1 2 3 4 5 6 7 8 9 10 11 12 The Mechanics of Nanoscale DevicesJohn Elie Sader The University of Melbourne (Maths & Stats) Start time: 3:15 pm Date: Friday 12 September 2008 Location: Theatre 1, ICT Bldg (Rm:205-Flr:2, 111 Barry Street, Carlton) Nanomechanical sensors are often used to measure environmental changes with extreme sensitivity. Controlling the effects of surfaces and fluid dissipation presents significant challenges to achieving the ultimate sensitivity in these devices. In this talk, I will give an overview of theoretical/experimental work we are undertaking to explore the underlying physical processes in these systems. The talk will be general and aimed at introducing some recent developments in the field of nanomechanical sensors. Approximating an epidemic process by a branching process A. D. Barbour University of Zuerich Start time: 1:15 pm Date: Friday 29 August 2008 Location: Russell Love Theatre, Richard Berry Bldg The growth of an epidemic process is often effectively approximated in its early stages by that of an associated branching process, an observation first formalized by Whittle (1955). The same ideas frequently form the basis for describing the local structure of random graphs, including those exhibiting `small world' or power law behaviour. In this talk, we discuss ways of formalizing the approximation, by using stochastic coupling methods. A (Gentle) Introduction to Regular Variation and Applications Daniel Tokarev MASCOS, University of Melbourne Start time: 3:15 pm Date: Wednesday 20 August 2008 Location: Chemical & Biomolecular Engineering Theatre, Room G20, Chemical Regular Variation theory, introduced by J. Karamata, is concerned with relating asymptotic behaviour of (real) functions and operators, such as Laplace transform, convolution and n-fold functional iterate. It has become an important tool in Probability, Analysis and other fields, and has greatly simplified and clarified much theory, such as in theorems of Abelian and Tauberian type and results on maxima and sums of independent and identically distributed random variables. This talk is intended to introduce the concept of regular variation and some fundamental results and applications in Probability. Diffusion Approximation for a Heavily Loaded Multi-User Wireless Communication System with Cooperation Professor Ruth Williams University of California, San Diego Start time: 3:15 pm Date: Friday 1 August 2008 Location: Theatre 1, ground floor of the ICT Building, 111 Barry St, The University of Melbourne We consider a model for a cellular wireless communication system in which data is transmitted to multiple users over a common channel. For information theoretic reasons, the rate of transmission over this channel can be enhanced by cooperation. Assuming a fixed channel and that the average arrival rate of data for each user is known, we consider a simple scheduling policy which exploits cooperation and which has been shown to be throughput-optimal under Markovian assumptions. As a measure of performance under this policy, we establish a heavy traffic diffusion approximation for the workload process. This diffusion process is a semimartingale reflecting Brownian motion (SRBM) living in the positive orthant of N-dimensional space (where N is the number of users). Nominally, this SRBM has one direction of reflection associated with each of the 2^{N}-1 boundary faces. However, we show that in fact only those directions associated with the (N-1)-dimensional boundary faces matter in the heavy traffic limit. Based on joint work with Sumit Bhardwaj. The pivot algorithm for self-avoiding walks Nathan Clisby MASCOS, The University of Melbourne Start time: 3:15 am Date: Friday 9 May 2008 Location: Theatre 1, ground floor of the ICT Building, 111 Barry St, The University of Melbourne The self-avoiding walk (SAW) is an important model in statistical mechanics, as it is a standard model in the study of critical phenomena (phase transitions) and in addition it accurately characterises the excluded volume effect of real polymers (long chain molecules). The pivot algorithm is a technique with a long history, and is an extremely powerful tool in the study of SAWs. For a number of important quantities (e.g. critical exponents) it is by far the most efficient known method of calculation. It works via a Markov chain where successive SAWs are generated by attempting to 'pivot' part of the walk by rotating or reflecting the walk around a randomly selected pivot point. I will explain how to implement the pivot algorithm and why it is so effective, and then describe my current research: by incorporating additional geometric information while running the Markov chain it is possible to dramatically improve the speed of the algorithm. |