2026-03-17

[thinhong]

Model the movement between hospital and community implicitly using rho (meaning a fraction of community join the FOI of hospital and vice versa)

Characteristics of the hospital includes:

• N_community >> N_hospital
• Effective contact rate hospital >> community
• Susceptible at the beginning of outbreak hospital > community
• rho (the fraction of community affect hospital and vice versa, we set the same value so that community and hospital size is constant throughout the outbreak)

To-do:

• Run all simulations using different combination over a range of a, b, c, rho
• Showing outbreak size (X) in different a, b, c, rho
• Model outbreak size as a function of a + b + c + rho using GAM and RF (showing that they are equal in 2 different methods) - surrogate modelling
• See serosurveillance (S_H0) vs vax coverage (S_H0 + S_C0) in predicting outbreak size (X): it depends on c (S_H0 = c*S_C0)
• Discover (LATER)

Note: Use tensor product smooths for GAM to incorporate interactions