Multitask learning and nonlinear optimal control of the COVID-19 outbreak: A geometric programming approach

Abstract

We propose a multitask learning approach to learn the parameters of a compartmental discrete-time epidemic model from various data sources and use it to design optimal control strategies of human-mobility restrictions that both curb the epidemic and minimize the economic costs associated with implementing non-pharmaceutical interventions. We develop an extension of the SEIR epidemic model that captures the effects of changes in human mobility on the spread of the disease. The parameters of the model are learned using a multitask learning approach that leverages both data on the number of deaths across a set of regions, and cellphone data on individuals’ mobility patterns specific to each region. Using this model, we propose a nonlinear optimal control problem aiming to find the optimal mobility-based intervention strategy that curbs the spread of the epidemic while obeying a budget on the economic cost incurred. We also show that the solution to this nonlinear optimal control problem can be efficiently found, in polynomial time, using tools from geometric programming. Furthermore, in the absence of a straightforward mapping from human mobility data to economic costs, we propose a practical method by which a budget on economic losses incurred may be chosen to eliminate excess deaths due to over-utilization of hospital resources. Our results are demonstrated with numerical simulations using real data from the COVID-19 pandemic in the Philadelphia metropolitan area.

Optimal spatial evaluation of a pro rata vaccine distribution rule for COVID-19

“…We leave for future work an investigation of the case where the number of individuals with vaccine failure is sufficiently large. Finally, we have only considered uncertainty that is either resolved before the allocation decision or never resolved at all; another possibility is that more information about parameters is learned during the allocation decision (see, for instance 59 , where cell phone data gradually informs mobility patterns for COVID-19). In such cases, adaptive management approaches may be of relevance 60 .…”

Ensemble Prediction Method Based on Decomposition–Reconstitution–Integration for COVID-19 Outbreak Prediction

“…These models, which rely on a single source of error, play a crucial role in the short-term forecasting of the COVID-19 epidemic. However, due to unstable incubation periods, asymptomatic patients, and epidemic prevention policies, there are complex time-series relationships in the epidemic time-series data, and the non-linearity and non-smoothness of the epidemic data become a great challenge for the data-driven class of epidemic prediction [14,15]. The present-day non-linear modeling techniques that are widely utilized in this domain include the likes of artificial neural networks (ANNs) [16,17], long short-term memory (LSTM) [18][19][20], and gate recurrent units (GRUs) [21].…”

Geometric programming solution of second degree difficulty for carbon ejection controlled reliable smart production system

“…Kalaiarasi et al [19] developed an economic quantity model with numerous items and solved their model with geometric programming in an uncertain manner using a variety of membership functions for the fuzzification process. Furthermore, Hayhoe et al [15] used a posynomial geometric programming approach to solve an optimal control problem and find strategies for the best control method.…”