MATRIX Research Program: Multivariate Dependence Modeling: Theory and Applications

Dependent data is widely observed across many different scientific disciplines, such as finance, agriculture, astronomy, hydrology and climatology, ecology and geology. They are also frequently used in health science and socio-economic disciplines. The recent development of powerful data gathering technology has brought larger and more complex datasets, which calls for new statistical tools developments.

The goal of this workshop is to bring together a large group of internationally recognised experts as well as early career researchers working in different areas of multivariate dependence modeling to exchange ideas. The workshop will also help consolidate collaborations for interdisciplinary researches.

The topics of interest to be discussed include recent advances in copula modeling and construction of flexible distributions for multivariate non-Gaussian data, with applications in actuarial and environmental science and finance, e.g., modeling a probability of default for a credit portfolio. Another important topic is modeling data with spatio-temporal dependence and time series data. Applications include modeling environmental data such as pollution levels in a certain geographical region, or predicting geohazard events, such as major landslides. A third topic deals with multivariate extremes, very rare events that affect large areas and may have a very significant impact on environment and economy, such as prolonged droughts or severe floods. Another very important topic is weather and climate modeling, including improved short-term and long-term forecasts. Other interesting topics include applications in astronomical sciences, variable selection methods and Bayesian methods for big data.

This MATRIX Research Program is partially supported by AMSI and AustMS through the AMSI-AustMS Workshop Funding.