• Corpus ID: 220936000

Interpretable Sequence Learning for COVID-19 Forecasting

  title={Interpretable Sequence Learning for COVID-19 Forecasting},
  author={Sercan {\"O}. Arik and Chun-Liang Li and Jinsung Yoon and Rajarishi S. Sinha and Arkady Epshteyn and Long T. Le and Vikas Menon and Shashank Singh and Leyou Zhang and Nate Yoder and Martin Nikoltchev and Yash Sonthalia and Hootan Nakhost and Elli Kanal and Tomas Pfister},
We propose a novel approach that integrates machine learning into compartmental disease modeling to predict the progression of COVID-19. Our model is explainable by design as it explicitly shows how different compartments evolve and it uses interpretable encoders to incorporate covariates and improve performance. Explainability is valuable to ensure that the model's forecasts are credible to epidemiologists and to instill confidence in end-users such as policy makers and healthcare institutions… 

Inter-Series Attention Model for COVID-19 Forecasting

A new neural forecasting model, called Attention Crossing Time Series (ACTS), that makes forecasts via comparing patterns across time series obtained from multiple regions that outperforms all the leading COVID-19 forecasters highlighted by CDC.

Interim evaluation of Google AI forecasting for COVID-19 compared with statistical forecasting by human intelligence in the first week

Google AI forecasting with a statistical model by human intelligence for COVID-19 outbreak shows Googles prediction has negatively correlated and greater difference with the data than the results, and the model is slightly correlated but not significant.

A Brief Summary on Covid-19 Pandemic & Machine Learning Approaches

This paper provides an overview of the machine learning approaches utilized in these pandemic-related problems including convolutional neural networks, reinforcement learning, and graph neural networks and discusses the possible adverse effects of utilization of certain machineLearning approaches in such a critical setup.

EINNs: Epidemiologically-Informed Neural Networks

This work introduces a new class of physics-informed neural networks-EINNcrafted for epidemic forecasting that allows neural models have the flexibility to learn the disease spread dynamics and use auxiliary features in a general framework.

SIMLR: Machine Learning inside the SIR model for COVID-19 Forecasting

The SIMLR model, which incorporates machine learning (ML) into the epidemiological SIR model, is applied to data from in Canada and the United States and it is shown that its mean average percentage error is as good as state-of-the-art forecasting models, with the added advantage of being an interpretable model.

Real-time and Explainable Detection of Epidemics with Global News Data

This study investigated how to detect and alert infectious diseases that might develop into pandemics soon, even before the information about a specific disease is aggre-gated, using only global news data, which are easily accessible in real time.

Deep Grey-Box Modeling With Adaptive Data-Driven Models Toward Trustworthy Estimation of Theory-Driven Models

A framework is presented that enables us to analyze a regularizers’ behavior empirically with a slight change in the neural net’s architecture and the training objective and it is shown that this approach can be trusted when not sure what regularizer is suitable for the given data.

Differentiable agent-based epidemiological modeling for end-to-end learning

G RAD ABM is introduced which is a new scalable, fast and differentiable design for ABMs that makes it amenable to be merged with deep neural networks and seamlessly integrate heterogeneous data sources to help with calibration, forecasting and policy evaluation.

Direct modelling from GPS data reveals daily-activity-dependency of effective reproduction number in COVID-19 pandemic

During the COVID-19 pandemic, governments faced difficulties in implementing mobility restriction measures, as no clear quantitative relationship between human mobility and infection spread in large



Curating a COVID-19 data repository and forecasting county-level death counts in the United States

This paper presents the continuous curation of a large data repository containing COVID-19 information from a range of sources, and develops and combines multiple forecasts using ensembling techniques, resulting in an ensemble that achieves a coverage rate of more than 94% when averaged across counties for predicting cumulative recorded death counts two weeks in the future.

Global prediction of unreported SARS-CoV2 infection from observed COVID-19 cases

Summary: Estimation of infectiousness and fatality of the SARS-CoV-2 virus in the COVID-19 global pandemic is complicated by ascertainment bias resulting from incomplete and non-representative

Accurate Statistics on COVID-19 Are Essential for Policy Guidance and Decisions.

Key epidemiological concepts are reviewed and some of the preventable methodologic errors that have arisen in reporting on the COVID-19 crisis are discussed.

Demographic science aids in understanding the spread and fatality rates of COVID-19

The role of age structure in deaths thus far in Italy and South Korea and how the pandemic could unfold in populations with similar population sizes but different age structures are examined, showing a dramatically higher burden of mortality in countries with older versus younger populations.

TDEFSI: Theory Guided Deep Learning Based Epidemic Forecasting with Synthetic Information

Theory Guided Deep Learning Based Epidemic Forecasting with Synthetic Information is presented, an epidemic forecasting framework that integrates the strengths of deep neural networks and high-resolution simulations of epidemic processes over networks and outperforms the other methods at the state and county levels in the USA.

Differential Effects of Intervention Timing on COVID-19 Spread in the United States

Counterfactual simulations indicate that, had these same control measures been implemented just 1-2 weeks earlier, a substantial number of cases and deaths could have been averted and underscore the importance of early intervention and aggressive response in controlling the COVID-19 pandemic.

An introduction to compartmental modeling for the budding infectious disease modeler

This work provides a basic introduction to disease modeling that is primarily intended for individuals who are new to developing SIR-type models, and highlights several common issues encountered when structuring and analyzing these models.

Contributions to the mathematical theory of epidemics—I

A mathematical investigation has been made of the progress of an epidemic in a homogeneous population, finding a threshold density of population is found to exist, which depends upon the infectivity, recovery and death rates peculiar to the epidemic.