ML APPLICATIONS

Neural Relational Autoregression for High-Resolution COVID-19 Forecasting

September 23, 2020

Abstract

Forecasting COVID-19 poses unique challenges due to the novelty of the disease, its unknown characteristics, and substantial but varying interventions to reduce its spread. To improve the quality and robustness of forecasts, we propose a new method which aims to disentangle region-specific factors -- such as demographics, enacted policies, and mobility -- from disease-inherent factors that influence its spread. For this purpose, we combine recurrent neural networks with a vector autoregressive model and train the joint model with a specific regularization scheme that increases the coupling between regions. This approach is akin to using Granger causality as a relational inductive bias and allows us to train high-resolution models by borrowing statistical strength across regions. In our experiments, we observe that our method achieves strong performance in predicting the spread of COVID-19 when compared to state-of-the-art forecasts.

Download the Paper

AUTHORS

Written by

Maximilian Nickel

Levent Sagun

Mark Ibrahim

Matt Le

Timothee Lacroix

Publisher

FB Data for Good

Related Publications

October 29, 2021

ML APPLICATIONS

Antipodes of Label Differential Privacy: PATE and ALIBI

Mani Malek, Ilya Mironov, Karthik Prasad, Igor Shilov, Florian Tramer

October 29, 2021

November 10, 2020

RESEARCH

ML APPLICATIONS

Adversarial Soft Advantage Fitting: Imitation Learning without Policy Optimization

Joelle Pineau, Christopher Pal, Derek Nowrouzezahrai, Julien Roy, Paul Barde, Wonseok Jeon

November 10, 2020

October 22, 2020

RESEARCH

SPEECH & AUDIO

Implicit Rank-Minimizing Autoencoder

Li Jing, Jure Zbontar, Yann LeCun

October 22, 2020

August 23, 2020

AR/VR

RESEARCH

Improving Vision-and-Language Navigation with Image-Text Pairs from the Web

Dhruv Batra, Devi Parikh, Arjun Majumdar, Ayush Shrivastava, Peter Anderson, Stefan Lee

August 23, 2020