A probabilistic infection model for efficient trace-prediction of disease outbreaks in contact networks

Willian Qian, Sanjukta Bhowmick, Marty O’Neill, Susie Ramisetty-Mikler, Armin R. Mikler

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We propose a novel method which we call the Probabilistic Infection Model (PIM). Instead of stochastically assigning exactly one state to each agent at a time, PIM tracks the likelihood of each agent being in a particular state. Thus, a particular agent can exist in multiple disease states concurrently. Our model gives an improved resolution of transitions between states, and allows for a more comprehensive view of outbreak dynamics at the individual level. Moreover, by using a probabilistic approach, our model gives a representative understanding of the overall trajectories of simulated outbreaks without the need for numerous (order of hundreds) of repeated Monte Carlo simulations. We simulate our model over a contact network constructed using registration data of university students. We model three diseases; measles and two strains of influenza. We compare the results obtained by PIM with those obtained by simulating stochastic SEIR models over the same the contact network. The results demonstrate that the PIM can successfully replicate the averaged results from numerous simulations of a stochastic model in a single deterministic simulation.

Original languageEnglish (US)
Title of host publicationComputational Science – ICCS 2020 - 20th International Conference, Proceedings
EditorsValeria V. Krzhizhanovskaya, Gábor Závodszky, Michael H. Lees, Peter M.A. Sloot, Peter M.A. Sloot, Peter M.A. Sloot, Jack J. Dongarra, Sérgio Brissos, João Teixeira
PublisherSpringer
Pages676-689
Number of pages14
ISBN (Print)9783030503703
DOIs
StatePublished - 2020
Event20th International Conference on Computational Science, ICCS 2020 - Amsterdam, Netherlands
Duration: Jun 3 2020Jun 5 2020

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12137 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference20th International Conference on Computational Science, ICCS 2020
CountryNetherlands
CityAmsterdam
Period6/3/206/5/20

Keywords

  • Computational epidemics
  • Outbreak simulation
  • SEIR model

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'A probabilistic infection model for efficient trace-prediction of disease outbreaks in contact networks'. Together they form a unique fingerprint.

  • Cite this

    Qian, W., Bhowmick, S., O’Neill, M., Ramisetty-Mikler, S., & Mikler, A. R. (2020). A probabilistic infection model for efficient trace-prediction of disease outbreaks in contact networks. In V. V. Krzhizhanovskaya, G. Závodszky, M. H. Lees, P. M. A. Sloot, P. M. A. Sloot, P. M. A. Sloot, J. J. Dongarra, S. Brissos, & J. Teixeira (Eds.), Computational Science – ICCS 2020 - 20th International Conference, Proceedings (pp. 676-689). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12137 LNCS). Springer. https://doi.org/10.1007/978-3-030-50371-0_50