SIR-Model
Control-Oriented Infection Networks Through a Linear-Systems Lens
A paper-backed research software project that treats infection spread as a systems problem: linear-state modeling, source localization, topology identification, and recovery design rather than generic epidemiology alone.
A Systems Project First
The value of this project is not just that it simulates infection spread. It reframes infection networks in a way that makes sensing, identification, and recovery design share one technical language.
Not generic epidemiology
The point of the project is not forecasting outbreaks at a dashboard level. It is to study infection spread as a networked dynamical system with identification and control structure.
Paper-backed and code-backed
The project is anchored by the Information Sciences paper and companion work on source localization and topology discovery, with code as a compact research reference rather than a productized toolkit.
Small but real
This page should read like a serious research-software note: focused scope, clear systems lens, and direct evidence links without inflated platform framing.
What Changed Once the Model Became a State System
The strongest part of the project is the modeling translation itself. Once the spread process is written in systems terms, multiple downstream questions become cleaner at the same time.
Source localization
Once the network is written as a linear system, source localization becomes much closer to an observability and sparse-recovery question than to a purely heuristic search.
Unknown infection time
Uncertain outbreak timing can be treated as an input-estimation problem, which makes the uncertainty legible instead of burying it inside ad hoc assumptions.
Topology discovery
Network recovery becomes sparse identification over the adjacency structure, with convex relaxations providing a more principled route than guess-and-check structure search.
Recovery actions
Interventions can be modeled as inputs, which opens the door to controllability analysis and minimum-energy recovery design rather than only passive diagnosis.
A Compact Model With Several Useful Questions Inside It
Localization from partial sensing
Reason about where the outbreak started when only a subset of nodes is observed.
Structure recovery
Infer communication or infection topology instead of assuming the graph is known.
Input-aware diagnosis
Handle unknown start times and uncertain excitation history with a systems-estimation view.
Recovery design
Move beyond analysis and ask what interventions drive the network back toward a safer state.
Where to Read the Actual Work
A shorter engineering summary of why the linear-systems view matters for diffusion, sensing, topology discovery, and recovery.
OpenThe lab thread that connects source localization and topology recovery into one compact research layer.
OpenA microscopic-view Infection model based on linear systems — Information Sciences, 2020.
OpenPublic repository for the SIR-Model project and supporting simulation work.
Open