A sequel to this tutorial was presented at AAMAS 2017.
A synthetic population is a set of synthetic people and households, each associated with demographic variables and typical daily activity sequences, located geographically. A synthetic population synthesizes data from a variety of sources into a common, person-centric, framework. The process preserves the confidentiality of the individuals in the original data sets, yet produces realistic attributes and demographics for the synthetic individuals.
The resulting model is a dynamic representation of human mobility and interaction over the course of a normative day. From this we can also induce a social contact network, which is an interaction-based graph whose vertices are synthetic people, labeled by their demographics, and whose edges represent estimated contacts, labeled by duration of contact and type of activity.
A synthetic population is specific to a geographic location because of its dependence on "contingent realities" for the area -- demographics of people who live there and the distribution of actual activity locations. It provides a plausible, bottom-up mechanism for generating large scale structure without making assumptions about hierarchies or other stylized properties.
Synthetic populations, and more broadly, synthetic information systems, have been used in a number of domains, including epidemiology, disaster response, infrastructure modeling, climate change, land use modeling, demography, and more.
The goals of this tutorial are:
- To present an overview the state of the art in synthetic populations
- Discuss how to synthesize populations, networks and more broadly information
- Discuss some of the challenges in the field
- Illustrate how advances in computation and data are useful in advancing the field.
- Provide the audience with an interesting topical area to work in.
- Describe open problems and challenges in the area
Samarth Swarup (firstname.lastname@example.org)
Madhav V. Marathe (email@example.com)