CBMS Conference: Bayesian Forecasting and Dynamic Models

University Of California-Santa Cruz, Santa Cruz CA

Investigators

Abstract

This award supports the 2020 NSF-CBMS conference "Bayesian forecasting and dynamic models" hosted by the Department of Statistics at the University of California Santa Cruz, during August 10-14, 2020. The conference will feature three principal lecturers that will deliver 10 main lectures. Professor Mike West from Duke University, who is a foundational researcher and a major reference in the field of Bayesian forecasting and dynamic models will deliver 7 main lectures. Professor Hedibert Lopes from Insper and Professor Raquel Prado from UCSC will deliver 3 main lectures. The conference will also feature a case-study session in a specific area of application to expose junior participants to the process of developing focused statistical tools for highly structured time series data. In addition, the conference will offer "hands-on" sessions on practical data analysis and a panel session with industry experts from companies in Northern California. This will provide participants additional exposure on how Bayesian forecasting and dynamic models are applied in practical non-academic settings. Established and junior researchers, postdoctoral fellows and students will have the opportunity to learn and discuss the major foundational ideas as well as recent and modern models and computing methods in the area of Bayesian time series and dynamic modeling. The conference aims to attract new researchers to this field. Furthermore, given the regional emphasis of the conference, it is expected that the conference will provide an important opportunity for strengthening links and collaborations between multiple groups in the Western United States. Adequate modeling and forecasting of temporal data, particularly in large-dimensional settings, is key in a wide range of applications. This area has defined a major research arena in the mathematical and statistical sciences for years and has also led to intense research activity in methodological, computational and applied areas where these methods are used. In particular, recent important research advances in this area have led to a massive body of literature that comprise new sophisticated models and methods for analysis and forecasting of time series data, as well as powerful computational tools and related software for inference and forecasting in an efficient manner. Exploring, understanding, and applying these models and tools can be a daunting task for newcomers, imposing a steep barrier into the field. This conference, along with the monograph derived from it will facilitate introduction to the area by providing a comprehensive review of Bayesian modeling and forecasting tools. For more information, please refer to the conference webpage: http://cbms.soe.ucsc.edu/2020/ This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.