Andreas Graefe
Macromedia University

Presentation: Forecasting the 2020 US presidential election

Andrew Harvey
Emeritus Professor of Econometrics, University of Cambridge

Presentation: Score-driven Time Series with applications in economics, finance and the environment

Rob Hyndman
Professor, Monash University

Presentation: Ten years of forecast reconciliation

Henrik Madsen
Professor, Technical University of Denmark
Department of Applied Mathematics and Computer Science

Presentation: Forecasting for the Green Transition

Esther Ruiz
Universidad Carlos III de Madrid

Presentation: Forecast Uncertainty and Resampling Techniques

Rafal Weron
Wroclaw University of Science and Technology, Poland

Presentation: Recent advances in electricity price forecasting: A 2020 perspective


Varunraj Valsaraj, Senior Manager, SAS Research and Development

Enhancing Shipment Forecast for CPG companies using Machine Learning and Demand Sensing

Edwin Ng, Senior Data Scientist, Uber

Fritz Obermeyer, Senior Research Scientist, Uber AI

Probabilistic Forecasting with Pyro Pyro is an open source library for Deep Probabilistic Programming, enabling Bayesian modeling and inference in the Python and PyTorch ecosystem. In this talk we will introduce probabilistic programming approaches to forecasting univariate and multivariate time series. We will show how to generalize classical forecasting methods by: hierarchically coupling multiple time series, using neural networks in models and inference, and parallelizing to leverage modern hardware.