Epidemic Models with Manual and Digital Contact Tracing
summary
A Markovian stochastic SIR on a large, homogeneously mixing population uses infection rate β, recovery rate γ and diagnosis rate δ; manual tracing reaches each past contact with probability p, digital tracing alerts fraction π of app users. Analysis replaces individuals by contact-traced components and derives a component-level branching process that yields an individual effective reproduction R_E. Results: manual tracing needs very high p to suppress outbreaks, digital needs very high π (≈0.9) to match, combining them is synergistic not multiplicative, and a counterintuitive rise in component reproduction with p appears while individual R_E nonetheless falls. Models omit delays and network structure; adding delays is work in progress.
AI-written summary · about
chapters (9)
- 0:00Welcome
- 6:32Epidemic modeling overview
- 12:27SIR model assumptions
- 20:09Manual contact tracing limits
- 29:22Digital tracing effectiveness
- 36:38Branching process analysis
- 56:29Reproduction number insights
- 1:06:37Future model improvements
- 1:07:44Q&A
related talks
14mA Social Network Under Social Distancing: Risk-Driven Backbone Management During COVID-19
46mA Framework for Imperfectly Observed Networks
37mCOVID-19 impact on RIPE Atlas #internet #measurements
21mFrom Reactive to Proactive: How AI is Redefining Resilience in Modern Networks
32mActive Fault Finding in Networks
35mTrying to De-Anonymize I2P Network Participants
31mCausalsim: Trace-Driven Simulation For Network Protocols
15mDecoding RSSI Compression in RFID: Dynamic RCS Modeling and Tag-Intrinsic Power Metrics..