CAST Seminar: Melody Ghahramani

Join us at the CANSSI Ontario STatistics Seminars (CAST) with

Melody Ghahramani

Professor
Department of Mathematics and Statistics
University of Winnipeg

Talk Title

On Time Series Regression Models for Zero-Inflated Proportions: parametric and semi-parametric approaches

Abstract

Time series of proportions are often encountered in applications such as ecology, environmental science and public health. Strategies for such data include linear regression after logistic transformation. Though easy to fit, the transformation approach renders covariate effects uninterpretable on the scale on which they were observed owing to Jensen’s inequality. An alternative to the transformation approach has been to directly model the response via the beta distribution. In this talk, we extend zero-inflated beta regression models for independent proportions to time series data that is bounded over the unit interval and that may take on zero values. Estimation is within the partial-likelihood framework and is computationally feasible to implement. We outline the asymptotic theory of our maximum partial likelihood estimators under mild regularity conditions and investigate their bias and variability using simulation studies. The utility of our method is illustrated using real data. In addition, a preliminary investigation of an approach robust to variance misspecification is presented. This work is joint with Antonio Axalan (a University of Winnipeg mathematics and statistics major graduate) and Dean Slonowsky (Vancouver Island University).

Speaker Profile

Melody Ghahramani is a professor of Statistics at the University of Winnipeg. She joined the Department of Mathematics and Statistics at the University of Winnipeg in 2007. Her early work focused on developing semi-parametric methods in the Godambe estimating function sense for time series data. A colleague sparked her interest in James-Stein shrinkage estimation with whom she develops shrinkage estimation for various time series models. She is passionate about training undergraduate students in statistics research and has co-authored statistics methods papers with several of them. On the applied side, she has been involved in projects on trend analysis of environmental sciences time series. Her research program has been funded by Discovery grants from the Natural Sciences and Engineering Research Council of Canada.

Her past service to the profession includes serving as board member of the Statistical Society of Canada (SSC), chair of the Statistics Education Committee of the SSC, treasurer of the Business and Industrial Statistics Section of the SSC, and member of the CANSSI Prairies advisory committee.

Melody completed a BSc in Mathematics and Statistics (First Class Honours) at the University of Manitoba, a Master’s in Statistics at Simon Fraser University and a PhD in Statistics at the University of Manitoba.