Statistical inference in realistic network models
Joost Jorritsma
Florence Nightingale Research Fellow
Department of Statistics
Interview: TU Eindhoven - Assistant professor in Statistics
What can we infer from a network snapshot?
Growing labeled network
1
2
3
4
5
6
7
8
9
10
Unlabeled snapshot
Estimate
Patient zero
Transmission mechanism
Structural features
Degrees, Correlations,
Geometry
Random graphs
Statistical modeling
Benchmark for inference
Inference: simplified versus realistic models
Well understood
Less understood
Trees
Dense graphs
Sparse graphs
Trees
Sparse graphs
Bringing structure and inference together
# Infections over time
Models
Geometry, Heterogeneity
Component structure
Spreading dynamics
Outbreak variability
Intervention modeling
Statistical inference
Arrival-time estimation
Hypothesis testing
1
2
3
4
5
6
7
8
9
10
Looking ahead
Veni: Criticality in random graphs
Probabilistic tools for structural analysis
Statistical inference in realistic networks
Estimation, testing, network archaeology
Robustness & differential privacy
Supervision & collaboration
Workshops & reading groups
PhD, MSc, BSc supervision
Industry experience (Air France – KLM)
Statistical inference in realistic network models
Joost Jorritsma
Florence Nightingale Research Fellow
Interview: TU Eindhoven - Assistant professor in Statistics