Seeking Significant Relationships
The information age has brought us a seemingly boundless wealth of data. The statistical analysis of these data enhances our ability to understand natural phenomena, anticipate economic movements and discover the causes of various illnesses.
Dr. Christian Genest, Canada Research Chair in Stochastic Dependence Modeling, is concerned with the development of statistical models and inference techniques in search of significant relationships between heterogeneous variables. Beyond the intricacies associated with the treatment of large data sets, this endeavour poses numerous methodological challenges related to the presence of sampling error and the possibility of rounding, censoring or truncating in the data. Time and various other covariates can also influence the behaviour of the variables of interest and must be taken into account in the study of dependence and its effects. Genest and his team are studying these issues, notably through the theory of copulas. Their work extends from model building to inference and prediction using methods based on multivariate analysis, nonparametric statistics, time series and the theory of empirical processes. Genest and his team develop new statistical techniques and, in collaboration with subject-matter specialists, adapt them to various contexts in insurance, finance and hydrology, thereby contributing to the advancement of theory and practice in those fields.