Modeling on social spread from immunity

Abstract

We are now planning to make a transmission model of infectious diseases in the scale of a city. People live in the city contacting other persons with daily life. The model regards a contact as a source of infection. A person will be simulated as a simple system of differential equations. As a candidate of differential equations, we are now investigating Marchuk's simple model. We adopt Marchuk's simple model because it has formation time, i.e., latent time. As Dr. Takeuchi showed, latent time is very important. There remain problems of choosing parameters for special diseases. We are now planning to use Marquardt method to minimize residuals form clinical data to estimate parameters. As for contacts, there are many approaches. The approach of the MIDAS project is very intensive. Our approach is simple. There are about 30,000 Japanese every fifteen minutes daily life data, sleeping, eating, work, study, house keeping, etc. Our approach is to make virtual families, husband, wife, children in a city and assign actions from the every fifteen minutes data statistically and estimate their contacts in the companies or schools, etc.