Past IMC Colloquium Abstracts
IMC Colloquium Series: Jump Detection with Wavelets for High-Frequency Financial Time Series
Ramazan Gençay
Department Of Economics, Simon Fraser University
Date: Apr 16, 2010
Time: 11:30 - 12:30
Room: ASB10900
WebSite: Click Here
Abstract
This paper introduces a new nonparametric test based on wavelets to detect jump arrival times in high frequency financial time series data. The asymptotic distribution of the test is derived. We demonstrate that the test is robust for different specifications of price processes and the presence of the microstructure noise. A Monte Carlo simulation is conducted to show that the test has good power and size. Further, we examine the multi-scale jump dynamics in U.S. equity markets. The main findings are as follows. First, the jump dynamics of equities are entirely different across different time scales, suggesting that choosing a proper sampling frequency is important to extract full jump dynamics. Second, although arrival densities of positive jumps and negative jumps are symmetric across different time scales, the magnitude of jumps is distributed asymmetrically at high frequencies. Third, only twenty percent of jumps occur in the trading session from 9:30AM to 4:00PM, suggesting that jumps are largely determined by news rather than liquidity shocks.
Video Archives
High Resolution: 
Low Resolution:
IMC Colloquium Series: "Modeling to Estimate Weight Gain from Energy Imbalances"
Diane Finegood and Penny Deck
Department of Biomedical Physiology and Kinesiology, Simon Fraser University
Date: Mar 19, 2010
Time: 10:30 - 12:30
Room: ASB10900
WebSite: Click Here
Abstract
Several model based approaches have been used to estimate the energy imbalance that accounts for population weight gain in recent decades. These estimates suggest that differences of as little as 50 to 100 calories excess energy intake over expenditure accounts for the rising prevalence of obesity. These estimates have led to the suggestion that small decreases in energy balance should accordingly result in weight loss. We will explore this concept and what population level and individual level models suggest are appropriate strategies to reverse the epidemic of obesity.
Video Archives
High Resolution:
Low Resolution:
IMC Colloquium Series: "From the Unbounded to the (Artificially) Bounded"
Nilima Nigam
SFU Department of Mathematics
Date: Mar 05, 2010
Time: 11:30 - 12:30
Room: ASB10900
WebSite: Click Here
Abstract
The study of wave interactions with obstacles has a rich mathematical history. Interesting and challenging issues arise when we numerically simulate these phenomena. For example, we can only discretize a bounded region, so one method is to introduce an artificial surface enclosing the obstacle. Instead of allowing waves to scatter off bounded obstacles in free space, we now have to study their action inside this computational box. What effect does the truncation have on the original model problem, which is posed in all of space? What further effects do discretization errors bring? We quickly survey some techniques for describing the so-called non-reflecting boundary conditions, and present some new work on the construction of exact Dirichlet-to-Neumann maps. Along the way, we need tools from functional analysis, PDE theory, numerical analysis and asymptotics.
IMC Colloquium Series: "Geographic Injury Surveillance"
Nadine Schuurman
Department of Geography, Simon Fraser University
Date: Feb 05, 2010
Time: 11:30 - 12:30
Room: ASB10900
WebSite: Click Here
Abstract
Each year, 5 million people die from injuries sustained from road traffic collisions, as a result of falls, on the, job, or due to violence. Another 100 million people are severely injured or permanently disabled by these mechanisms. Together, these events, which are often preventable, are a leading cause of potential years of life lost, lost productivity, and inestimable hardships for individuals, families and society. In British Columbia, more than 400,000 are injured each year. Of these, 26,000 require hospitalization, over 9,000 are left with permanent disabilities, and 1600 die, resulting in the loss of over 37,000 potential years of life. In Canada and around the world, injury is increasingly recognized as an important public health issue requiring investment in research, thoughtful analysis and vigorous action. A public health approach - based on identifying and reducing risk - to injury control is considered to be most effective method of reducing the burden of injury. This talk focuses on way that geographic information science, spatial analysis and cartography can be used to identify clusters of injury, correlates with socio-economic status, impediments to access, as well as the role of enhanced data collection in resource poor environments. The role of GIS in understanding and mitigating injury will be illustrated by reviewing multiple injury surveillance projects in British Columbia, Canada and South Africa.
Video Archives
High Resolution:
Low Resolution:
IMC Colloquium Series: "Engineering Security: A Formal Approach"
Uwe Glasser
SFU IRMACS Centre and School of Computing Science
Date: Jan 22, 2010
Time: 12:30 - 13:30
Room: ASB10900
WebSite: Click Here
Abstract
Security builds on abstractions that call for accurate models to be analyzed and validated so as to ensure they serve their purpose. Abstract operational security requirements can naturally be described in terms of discrete mathematics and computational logic. Mathematical precision is essential for reasoning about critical properties and to uncover and eliminate design flaws and weaknesses that often go unnoticed otherwise; it is also a prerequisite for using computational methods and tools, for instance, in experimental studies by means of computer simulation as a feasible alternative to performing real world experiments. We illustrate the approach by presenting our experience with various collaborative R&D projects, with government agencies and industrial partners, on computational criminology, decision support for coastal surveillance, identity management, and situation analysis.
Video Archives
High Resolution:
Low Resolution:
IMC Colloquium Series: "A Particle Filter Approach to Identification of Discrete Stochastic Nonlinear Processes Under Missing Observations"
Bhushan Gopaluni
Department Of Chemical & Biological Engineering, University Of British Columbia
Date: Nov 27, 2009
Time: 11:30 - 12:30
Room: ASB10900
WebSite: Click Here
Abstract
A novel maximum likelihood solution to the problem of identifying parameters of a nonlinear model under missing observations will be presented. If the observations are missing, then it is difficult to build a partial likelihood function consisting of only the available observations. Hence, a variant of expectation maximization (EM) algorithm, which uses the expected value of the complete log-likelihood function including the missing observations, is developed. The expected value of the complete log-likelihood (E-step) in the EM algorithm is approximated using particle filters and smoothers. New expressions for particle filters and smoothers under missing observations are derived. In order to reduce the variance on the smoothed states, a point-wise (as opposed to path- based) state estimation procedure is used. The maximization step (M-step) in the EM algorithm is performed using standard optimization routines. The proposed nonlinear identification approach is illustrated through numerical and industrial examples.
Video Archives
High Resolution:
Low Resolution:
IMC Colloquium Series: "Measuring Social Influences on Health"
Brian Krauth
Department of Economics, SFU
Date: Nov 13, 2009
Time: 11:30 - 12:30
Room: ASB10900
WebSite: Click Here
Abstract
Conventional wisdom and substantial evidence suggest that a person's social environment affects both health behavior and health outcomes. However, obtaining credible measurements of these effects runs into the problem that social environment is at least partly chosen. As a result, the set of likely omitted/confounding variables is virtually unlimited. My talk will describe a research program that aims to quantify the uncertainty associated with effect measurements in this context. The unifying idea is to characterize the importance of omitted variables relative to included control variables. I will also describe applications to the measurement of peer effects in youth smoking, and of the effect of income inequality on health.
Video Archives
High Resolution:
Low Resolution:
IMC Colloquium Series: "Recent Advances in Medical Imaging"
Joe Qranfal
Department of Mathematics, SFU
Date: Oct 30, 2009
Time: 11:30 - 12:30
Room: ASB10900
Abstract
Images and visualization have become increasingly important in many areas of science and technology. Advances in hardware and software have allowed computerized image processing to become a standard tool in many scientific applications, including medical imaging. In this talk, we see how we model and solve the inverse problem of reconstructing a dynamic medical image where the signal strength changes substantially over the time required for data acquisition. We use a stochastic approach based on a Markov process to model the problem. We introduce a novel proximal approach and apply it during the Kalman filter algorithm to ensure positivity and spatial regularization. We test our method for the case of image reconstruction in time-dependent single photon emission computed tomography (SPECT). Numerical results corroborate the effectiveness of our approach.
Video Archives
High Resolution:
Low Resolution:
IMC Colloquium Series: "Simple questions with difficult answers: Modelling the HIV epidemic in BC"
Robert Hogg
Faculty of Health Sciences, SFU
Date: Oct 16, 2009
Time: 11:30 - 12:30
Room: ASB10900
WebSite: Click Here
Abstract
TBA
Video Archives
High Resolution:
Low Resolution:
IMC Colloquium Series: "The Arithmetic of Sentencing"
Pierre Tremblay
School of Criminology, University Of Montreal
Date: Oct 02, 2009
Time: 11:30 - 12:30
Room: ASB10900
Abstract
I will talk about scaling of severity of various penalties (fine, probation, prison)our ability to establish reasonable severity equivalencies. The main subject matter of the talk will be to ask if lawyers and judges have a more cogent understand of penal metrics than the general public.
Video Archives
High Resolution:
Low Resolution:
IMC Colloquium Series: "Systems Thinking for Health Systems Renewal"
Allan Best
Centre for Clinical Epidemiology and Evaluation, Vancouver General Hospital
Date: Sep 18, 2009
Time: 11:30 - 12:30
Room: ASB10900
WebSite: Click Here
Abstract
Health policy and health organisation leaders are emphasizing the importance of evidence when searching for new ways to tackle complex public health issues (such as obesity and health inequalities) and address enduring organisational challenges (such as safety, efficiency and patient-centredness). Yet, how we think about the challenge of moving from evidence to policy, or from evidence to organisational change, profoundly shapes the ways we work and the results that we get. Today, our ways of thinking are changing fundamentally, as we come to understand that our health systems are complex systems, demanding fresh conceptual models, more integrative research methodology, and organizational forms that nurture and support evidence-informed decision making. We need more interdisciplinarity, more systems thinking, more problem-based research, and some serious restructuring of the ways in which universities, research funders, and health service organizations are designed. If this systemic change is stifled, we risk never realizing our potential for creating and using knowledge in a collaborative way, and we commit an injustice to the populations we serve. Imagine a system in which researchers, decision makers and practitioners work collaboratively throughout the full knowledge creation, synthesis, and application process. Think about how tenure and promotion policies might change if contributions to social good were valued as much as contributions to knowledge. Reflect on how knowledge from randomization and experimental control might be balanced by real-world data and learning from practice. Ask what it will take to develop new analytic methods that capture the dynamic, multilevel nature of change as health systems work to continuously improve.
