<span style="font-variant: small-caps">EpiInfer</span>: A Non-Markovian Method and System to Forecast Infection Rates in Epidemics

Consider an evolving epidemic in which each person is either (S) susceptible and healthy; (E) exposed, contagious but asymptomatic; (I) infected, symptomatic, and quarantined; or (R) recovered, healthy, and susceptible. The inference problem, given (i) who is showing symptoms (I) and who is not (S,...

Full description

Saved in:
Bibliographic Details
Main Authors: Jovan Kascelan, Ruoxi Yang, Dennis Shasha
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Algorithms
Subjects:
Online Access:https://www.mdpi.com/1999-4893/18/7/450
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849714294696968192
author Jovan Kascelan
Ruoxi Yang
Dennis Shasha
author_facet Jovan Kascelan
Ruoxi Yang
Dennis Shasha
author_sort Jovan Kascelan
collection DOAJ
description Consider an evolving epidemic in which each person is either (S) susceptible and healthy; (E) exposed, contagious but asymptomatic; (I) infected, symptomatic, and quarantined; or (R) recovered, healthy, and susceptible. The inference problem, given (i) who is showing symptoms (I) and who is not (S, E, R) and (ii) the distribution of meetings among people each day, is to predict the number of infected people (state I) in future days (e.g., 1 through 20 days out into the future) for the purpose of planning resources (e.g., needles, medicine, staffing) and policy responses (e.g., masking). Each prediction horizon has different uses. For example, staffing may require forecasts of only a few days, while logistics (i.e., which supplies to order) may require a two- or three-week horizon. Our algorithm and system <span style="font-variant: small-caps;">EpiInfer</span> is a non-Markovian approach to forecasting infection rates. It is non-Markovian because it looks at infection rates over the past several days in order to make predictions about the future. In addition, it makes use of the following information: (i) the distribution of the number of meetings per person and (ii) the transition probabilities between states and uses those estimates to forecast future infection rates. In both simulated and real data, <span style="font-variant: small-caps;">EpiInfer</span> performs better than the standard (in epidemiology) differential equation approaches as well as general-purpose neural network approaches. Compared to ARIMA, <span style="font-variant: small-caps;">EpiInfer</span> is better starting with 6-day forecasts, while ARIMA is better for shorter forecast horizons. In fact, our operational recommendation would be to use ARIMA (1,1,1) for short predictions (5 days or less) and then <span style="font-variant: small-caps;">EpiInfer</span> thereafter. Doing so would reduce relative Root Mean Squared Error (RMSE) over any state of the art method by up to a factor of 4. Predictions of this accuracy could be useful for people, supply, and policy planning.
format Article
id doaj-art-4ada4d6facdc42469e374d13d93edca9
institution DOAJ
issn 1999-4893
language English
publishDate 2025-07-01
publisher MDPI AG
record_format Article
series Algorithms
spelling doaj-art-4ada4d6facdc42469e374d13d93edca92025-08-20T03:13:44ZengMDPI AGAlgorithms1999-48932025-07-0118745010.3390/a18070450<span style="font-variant: small-caps">EpiInfer</span>: A Non-Markovian Method and System to Forecast Infection Rates in EpidemicsJovan Kascelan0Ruoxi Yang1Dennis Shasha2Department of Computer Science, New York University Abu Dhabi, Abu Dhabi P.O. Box 129188, United Arab EmiratesDepartment of Biostatistics, School of Global Public Health, New York University, New York, NY 10012, USAComputer Science Department, Courant Institute of Mathematical Sciences, New York University, New York, NY 10012, USAConsider an evolving epidemic in which each person is either (S) susceptible and healthy; (E) exposed, contagious but asymptomatic; (I) infected, symptomatic, and quarantined; or (R) recovered, healthy, and susceptible. The inference problem, given (i) who is showing symptoms (I) and who is not (S, E, R) and (ii) the distribution of meetings among people each day, is to predict the number of infected people (state I) in future days (e.g., 1 through 20 days out into the future) for the purpose of planning resources (e.g., needles, medicine, staffing) and policy responses (e.g., masking). Each prediction horizon has different uses. For example, staffing may require forecasts of only a few days, while logistics (i.e., which supplies to order) may require a two- or three-week horizon. Our algorithm and system <span style="font-variant: small-caps;">EpiInfer</span> is a non-Markovian approach to forecasting infection rates. It is non-Markovian because it looks at infection rates over the past several days in order to make predictions about the future. In addition, it makes use of the following information: (i) the distribution of the number of meetings per person and (ii) the transition probabilities between states and uses those estimates to forecast future infection rates. In both simulated and real data, <span style="font-variant: small-caps;">EpiInfer</span> performs better than the standard (in epidemiology) differential equation approaches as well as general-purpose neural network approaches. Compared to ARIMA, <span style="font-variant: small-caps;">EpiInfer</span> is better starting with 6-day forecasts, while ARIMA is better for shorter forecast horizons. In fact, our operational recommendation would be to use ARIMA (1,1,1) for short predictions (5 days or less) and then <span style="font-variant: small-caps;">EpiInfer</span> thereafter. Doing so would reduce relative Root Mean Squared Error (RMSE) over any state of the art method by up to a factor of 4. Predictions of this accuracy could be useful for people, supply, and policy planning.https://www.mdpi.com/1999-4893/18/7/450epidemic modelingcompartmental modelsSEIRSforecastingcontact networksparameter estimation
spellingShingle Jovan Kascelan
Ruoxi Yang
Dennis Shasha
<span style="font-variant: small-caps">EpiInfer</span>: A Non-Markovian Method and System to Forecast Infection Rates in Epidemics
Algorithms
epidemic modeling
compartmental models
SEIRS
forecasting
contact networks
parameter estimation
title <span style="font-variant: small-caps">EpiInfer</span>: A Non-Markovian Method and System to Forecast Infection Rates in Epidemics
title_full <span style="font-variant: small-caps">EpiInfer</span>: A Non-Markovian Method and System to Forecast Infection Rates in Epidemics
title_fullStr <span style="font-variant: small-caps">EpiInfer</span>: A Non-Markovian Method and System to Forecast Infection Rates in Epidemics
title_full_unstemmed <span style="font-variant: small-caps">EpiInfer</span>: A Non-Markovian Method and System to Forecast Infection Rates in Epidemics
title_short <span style="font-variant: small-caps">EpiInfer</span>: A Non-Markovian Method and System to Forecast Infection Rates in Epidemics
title_sort span style font variant small caps epiinfer span a non markovian method and system to forecast infection rates in epidemics
topic epidemic modeling
compartmental models
SEIRS
forecasting
contact networks
parameter estimation
url https://www.mdpi.com/1999-4893/18/7/450
work_keys_str_mv AT jovankascelan spanstylefontvariantsmallcapsepiinferspananonmarkovianmethodandsystemtoforecastinfectionratesinepidemics
AT ruoxiyang spanstylefontvariantsmallcapsepiinferspananonmarkovianmethodandsystemtoforecastinfectionratesinepidemics
AT dennisshasha spanstylefontvariantsmallcapsepiinferspananonmarkovianmethodandsystemtoforecastinfectionratesinepidemics