COVID-19 dynamics across the US: A deep learning study of human mobility and social behavior

Bhouri, Mohamed Aziz; Costabal, Francisco Sahli; Wang, Hanwen; Linka, Kevin; Peirlinck, Mathias; Kuhl, Ellen; Perdikaris, Paris

doi:10.1016/j.cma.2021.113891

Cited by 44 publications

(47 citation statements)

References 40 publications

(32 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The ongoing COVID-19 pandemic has caused immense disruptions to our day-to-day normal life. Admittedly, human mobility plays an important role in spreading the virus ( Bhouri et al, 2021 , Bryant and Elofsson, 2020 , Chang et al, 2021 ). For this reason, to control the spread of the virus various human mobility restrictions, such as stay at home order, limited public gathering, non-essential business closures had been extensively exercised as a crucial non-pharmaceutical intervention tool all over the world.…”

Section: Introductionmentioning

confidence: 99%

Structural modeling of COVID-19 spread in relation to human mobility

Rafiq

Ahmed

Uddin

2022

Transportation Research Interdisciplinary Perspectives

View full text Add to dashboard Cite

Human mobility is considered as one of the prominent non-pharmaceutical interventions to control the spread of the pandemic (positive effect from mobility to infection). Conversely, the spread of the pandemic triggered massive changes to people’s daily schedules by limiting their movement (negative effect from infection to mobility). The purpose of this study is to investigate this bi-directional relationship between human mobility and COVID-19 spread across U.S. counties during the early phase of the pandemic when infection rates were stabilizing and activity-travel behavior reflected a fairly steady return to normal following the drastic changes observed during the pandemic’s initial shock. In particular, we applied Structural Regression (SR) model to investigate a bi-directional relationship between COVID-19 infection rate and the degree of human mobility in a county in association with socio-demographic and location characteristics of that county, and state-wide COVID-19 policies. Combining U.S. county-level cross-sectional data from multiple sources, our model results suggested that during the study period, human mobility and infection rate in a county both influenced each other, but in an opposite direction. Metropolitan counties experienced higher infection and lower mobility than non-metropolitan counties in the early stage of the pandemic. Counties with highly infected neighboring counties and more external trips had a higher infection rate. During the study period, community mitigation strategies, such as stay at home order, emergency declaration, and non-essential business closure significantly reduced mobility whereas public mask mandate significantly reduced infection rates. The findings of this study will provide important insights to policy makers in understanding the two-way relationship between human mobility and COVID-19 spread and to derive mobility-driven policy actions accordingly.

show abstract

Section: Introductionmentioning

confidence: 99%

Structural modeling of COVID-19 spread in relation to human mobility

Rafiq

Ahmed

Uddin

2022

Transportation Research Interdisciplinary Perspectives

View full text Add to dashboard Cite

show abstract

“…Using a LSTM network, for example, the combination of mobility and meteorological data was found to be the primary factor in case prediction [19]. A susceptible to infectious transition rate, as a function of mobility and social behavior with time-dynamics parameters, was modeled by a deep LSTM, which was further integrated into a susceptibleexposed-infectious-recovered (SEIR) model for case forecast [20]. A graph neural network is another attempt to capture the spatio-temporal dynamics, where spatial edges represent mobility-based inter-region connectivity and temporal edges represent node features through time [21].…”

Section: Existing Research On Mobility Patterns and Covid-19mentioning

confidence: 99%

Autoregressive count data modeling on mobility patterns to predict cases of COVID-19 infection

Zhao

Han

Wang

et al. 2022

Preprint

View full text Add to dashboard Cite

At the beginning of 2022 the global daily count of new cases of COVID-19 exceeded 3.2 million, a tripling of the historical peak value reported between the initial outbreak of the pandemic and the end of 2021. Aerosol transmission through interpersonal contact is the main cause of the disease’s spread, although control measures have been put in place to reduce contact opportunities. Mobility pattern is a basic mechanism for understanding how people gather at a location and how long they stay there. Due to the inherent dependencies in disease transmission, models for associating mobility data with confirmed cases need to be individually designed for different regions and time periods. In this paper, we propose an autoregressive count data model under the framework of a generalized linear model to illustrate a process of model specification and selection. By evaluating a 14-day-ahead prediction from Sweden, the results showed that for a dense population region, mobility data with a lag of 8 days is most reliable way of predicting the number of confirmed cases in relative numbers at a high coverage rate. It is sufficient for both of the autoregressive terms, studied variable and conditional expectation, to take one day back. For sparsely populated regions, a lag of 10 days produced the lowest error in absolute value for the predictions, where weekly periodicity on the studied variable is recommended for use. Interventions were further included to identify the most relevant mobility categories. Statistical features were also presented to verify the model assumptions.

show abstract

“…We consider a time window of 15 days. • DeepCovid-19 [24]: This is an existing deep learningbased model for COVID forecasting, that was designed for county level forecasting of positive cases using mobility features. Similar to LSTM, we use our ML-Dataset as a mobility feature.…”

Section: Baseline Modelsmentioning

confidence: 99%

“…Bhouri et al [24] used Google's mobility data to use as features for their proposed DL-based model for doing countywise forecasting of positive cases for over 200 counties. However, they do not address the correlation among different time series or the spatial dependency among different counties.…”

Section: Introductionmentioning

confidence: 99%

Deep diffusion-based forecasting of COVID-19 by incorporating network-level mobility information

Roy¹,

Sarkar²,

Biswas³

et al. 2021

Preprint

View full text Add to dashboard Cite

Modeling the spatiotemporal nature of the spread of infectious diseases can provide useful intuition in understanding the time-varying aspect of the disease spread and the underlying complex spatial dependency observed in people's mobility patterns. Besides, the county level multiple related time series information can be leveraged to make a forecast on an individual time series. Adding to this challenge is the fact that real-time data often deviates from the unimodal Gaussian distribution assumption and may show some complex mixed patterns. Motivated by this, we develop a deep learning-based time-series model for probabilistic forecasting called Auto-regressive Mixed Density Dynamic Diffusion Network (ARM3Dnet), which considers both people's mobility and disease spread as a diffusion process on a dynamic directed graph. The Gaussian Mixture Model layer is implemented to consider the multimodal nature of the realtime data while learning from multiple related time series. We show that our model, when trained with the best combination of dynamic covariate features and mixture components, can outperform both traditional statistical and deep learning models in forecasting the number of Covid-19 deaths and cases at the county level in the United States.

show abstract

COVID-19 dynamics across the US: A deep learning study of human mobility and social behavior

Cited by 44 publications

References 40 publications

Structural modeling of COVID-19 spread in relation to human mobility

Structural modeling of COVID-19 spread in relation to human mobility

Autoregressive count data modeling on mobility patterns to predict cases of COVID-19 infection

Deep diffusion-based forecasting of COVID-19 by incorporating network-level mobility information

Contact Info

Product

Resources

About