When Grace Tompkins attended Waterloo’s Grad Visit Day in 2019, she was presented with an opportunity to combine her interests in public health with her research in mathematics. The Master of Mathematics in Biostatistics program was the perfect blend of mathematical theory and epidemiology that she was looking for.
Longevity risk is an essential concept for life insurance companies and reinsurance companies. A deep understanding of the risk is crucial, as an underestimation of longevity can seriously affect the financial healthiness of pension plans and annuity funds.
Network Response Regression for Modeling Population of Networks with Covariates
Multiple-network data are fast emerging in recent years, where a separate network over a common set of nodes is measured for each individual subject, along with rich subject covariates information. Existing network analysis methods have primarily focused on modeling a single network, and are not directly applicable to multiple networks with subject covariates.
In this talk, we present a new network response regression model, where the observed networks are treated as matrix-valued responses, and the individual covariates as predictors. The new model characterizes the population-level connectivity pattern through a low-rank intercept matrix, and the parsimonious effects of subject covariates on the network through a sparse slope tensor. We formulate the parameter estimation as a non-convex optimization problem, and develop an efficient alternating gradient descent algorithm. We establish the non-asymptotic error bound for the actual estimator from our optimization algorithm. Built upon this error bound, we derive the strong consistency for network community recovery, as well as the edge selection consistency. We demonstrate the efficacy of our method through intensive simulations and two brain connectivity studies.
Join Zoom Meeting
Meeting ID: 844 283 6948
Passcode: 318995
Associate Professor Marius Hofert was awarded a Natural Sciences and Engineering Council of Canada (NSERC) Discovery Accelerator Supplement for his suggested new work on copula modeling with generative neural networks.
A statistics professor used his expertise in calculating probabilities to come up with a 98 winning percentage for Tim Hortons popular Roll up the Rim contest.Statistics and Actuarial Science PhD candidate Yilin Chen is one of two students to claim the 2020 Huawei Prize for Best Research paper by a Mathematics Graduate Student. The $4,000 prize affirms the value of Chen’s efforts to establish a framework for analyzing nonprobability survey samples in her winning paper: Doubly Robust Interference with Nonprobability Survey Samples.
They won’t graduate until next spring, but their startup has already taken flight.
The five founders of Gooloo recently earned a spot in the University of Waterloo’s Concept $5K Finals, where they explained their business model to a panel of judges alongside nine other student teams.
Four graduate students were awarded a departmental research presentation award by the Department of Statistics and Actuarial Science, but that's not all they have in common. They all came to Waterloo because they knew of the excellence of the Statistics programs, research, and professors. Their backgrounds vary, as do their research areas, but they have all had a great experience.
The University of Waterloo, which is among the top universities in the world for actuarial science and number one in co-operative learning, has combined these two things to help build a stronger insurance and pension industry in Indonesia.
A Machine Learning Approach to Portfolio Risk Management
Risk measurement, valuation and hedging form an integral task in portfolio risk management for insurance companies and other financial institutions. Portfolio risk arises because the values of constituent assets and liabilities change over time in response to changes in the underlying risk factors. The quantification of this risk requires modeling the dynamic portfolio value process. This boils down to compute conditional expectations of future cash flows over long time horizons, e.g., up to 40 years and beyond, which is computationally challenging.
This lecture presents a framework for dynamic portfolio risk management in discrete time building on machine learning theory. We learn the replicating martingale of the portfolio from a finite sample of its terminal cumulative cash flow. The learned replicating martingale is in closed form thanks to a suitable choice of the reproducing kernel Hilbert space. We develop an asymptotic theory and prove
convergence and a central limit theorem. We also derive finite sample error bounds and concentration inequalities. As application we compute the value at risk and expected shortfall of the one-year loss of some stylized portfolios.