SIS model

SIS model
Name	SIS model
Discipline	Epidemiology, Mathematical Biology, Network Science
Introduced	1920s–1930s
Key people	Andrey Kolmogorov, Rudolf Kermack, Andreas McKendrick, William O. Kermack
Methods	Compartmental modeling, Ordinary differential equations, Stochastic processes
Applications	Infectious disease modeling, Information diffusion, Ecology, Computer viruses

SIS model The SIS model is a compartmental infectious-disease model that classifies individuals into two epidemiological states and describes transitions between them using mathematical and computational tools. Developed within the tradition of early 20th-century mathematical epidemiology, the SIS framework underpins analyses in public health, theoretical ecology, and network science and has been refined by researchers across London School of Hygiene and Tropical Medicine, University of Cambridge, University of Oxford, Princeton University, and Stanford University.

Introduction

The SIS model divides a population into susceptible and infectious compartments and allows recovered individuals to return immediately to susceptibility, capturing diseases with no lasting immunity. Foundational contributors include Rudolf Kermack, Andreas McKendrick, and contemporaries such as Andrey Kolmogorov and William O. Kermack, while later formalizations connect to work at Bell Labs, Los Alamos National Laboratory, and institutions like Imperial College London. The model interfaces with concepts developed by Ronald Ross, Joseph L. Doob, Harold W. Kuhn, and has been applied in studies by groups at Centers for Disease Control and Prevention, World Health Organization, and regional public-health agencies.

Mathematical Formulation

The canonical SIS model is typically presented as a system of ordinary differential equations on a closed, homogeneously mixing population. Let S(t) and I(t) denote compartment sizes; parameters include transmission rate β and recovery rate γ, yielding dI/dt = β I (N − I)/N − γ I when total population N is constant. This deterministic formulation parallels early stochastic treatments influenced by Andrey Kolmogorov and later networked generalizations inspired by work at Santa Fe Institute and Massachusetts Institute of Technology. Matrix- and operator-based extensions employ adjacency matrices used in analyses from National Institute of Standards and Technology collaborations and linearization techniques akin to those in studies from California Institute of Technology.

Analysis and Properties

Equilibrium analysis identifies disease-free and endemic steady states; stability is governed by the basic reproduction number R0 = β/γ in homogeneously mixed settings. Bifurcation and stability results connect to mathematical tools developed by researchers at Institut des Hautes Études Scientifiques, Courant Institute, and ETH Zurich. Spectral properties on networks relate to results by scholars at University of Michigan and University of California, Los Angeles using the largest eigenvalue of contact matrices, drawing on techniques from Princeton Plasma Physics Laboratory and Max Planck Institute for Dynamics and Self-Organization. Stochastic versions introduce extinction probabilities and mean time to extinction, topics studied by teams at Los Alamos National Laboratory, RAND Corporation, and Imperial College London.

Extensions and Variants

Numerous extensions enrich the basic SIS framework: age-structured SIS models influenced by demographic methods from London School of Hygiene and Tropical Medicine, spatial SIS partial differential equation models developed at University of Paris, and network SIS models that incorporate heterogeneous contact patterns explored at Cornell University and University of Cambridge. Metapopulation and multi-strain SIS variants relate to work at Johns Hopkins University and University of California, Berkeley, while adaptive and temporal-network SIS models echo research from Santa Fe Institute and Max Planck Institute for Mathematics in the Sciences. Control-theoretic formulations use methods from Massachusetts Institute of Technology and ETH Zurich to design intervention strategies.

Applications and Case Studies

The SIS model serves as a minimal model for diseases such as gonorrhea, chlamydia, and some forms of common cold where immunity wanes quickly or is absent. Public-health analyses at Centers for Disease Control and Prevention and outbreak investigations by teams at World Health Organization have used SIS-inspired reasoning for settings with rapid reinfection. Beyond infectious disease, SIS frameworks describe information spread in online platforms studied by researchers at Facebook, Google, and Twitter (now X), and model computer-virus propagation in cybersecurity work at Microsoft Research and IBM Research. Ecological and evolutionary case studies at University of California, Davis and University of British Columbia employ SIS dynamics for host–pathogen systems and parasite transmission.

Numerical Methods and Simulation

Numerical treatment of SIS models ranges from classical explicit and implicit integrators for deterministic ordinary differential equations to Gillespie-style algorithms for exact stochastic simulation introduced by groups at Los Alamos National Laboratory and refined at Weizmann Institute of Science. Network-based simulations use agent-based frameworks developed at Sandia National Laboratories, Argonne National Laboratory, and in open-source platforms maintained by contributors from University of Washington and University of Illinois Urbana-Champaign. Parameter inference and uncertainty quantification leverage techniques from Columbia University, Harvard University, and University of Toronto including likelihood-based estimation, Bayesian computation, and ensemble Kalman filters.

Category:Mathematical models in epidemiology