22nd Oct: worshop on "Synthetic data meets simulation"

Synthetic data meets simulation workshop
Date: 22nd October 2010 10AM - 4PM
Venue: University Manchester

Sponsored by SIMIAN and CCSR (http://www.ccsr.ac.uk)

A workshop to explore synergies between synthetic data generation and agent-based modelling

Agent based modelling and synthetic data generation both produce artificial data based on theories or models. Synthetic data generation produces simulated population data usually from real data samples and/or data restricted in some way for confidentiality reasons. Agent based modelling generates data about simulated populations and could also be seen as a means of producing synthetic data.

This workshop has been conceived to explore whether research in both fields could be mutually informative and whether the techniques, methods and data products of each could enhance the other. Some ideas that may be worth exploring include: the use of synthetic data to assign properties to agents so that they possess the multivariate properties of the population of interest; and the use of agent based models to enhance longitudinal aspects of synthetic data (which may well have been generated cross-sectionally).

The workshop is free to attend. Lunch and refreshments will be provided. However places are limited, so please book early.

  • Jerry Reiter is Professor of Statistics Science at Duke University. He has is one of the world leading researchers on synthetic data generation, with particular reference to its applications to data privacy. More generally, he is interested in research involving survey methodology, especially dealing with missing data in complex surveys.
  • Mark Elliot has worked at the Centre for Census and Survey Research(CCSR) since 1996, mainly in the field of statistical confidentiality, founding the international recognised Confidentiality and Privacy Research Group (CAPRI) in 2002, and has managed numerous research projects within CAPRI remit. In 2005 he became Director of CCSR. He is one of the key international researchers in the field of Statistical Disclosure and has an extensive portfolio of research grants and publications in the field.
  • Bruce Edmonds is the Director of the Centre for Policy Modelling, and a Senior Research Fellow at the Manchester Metropolitan Unviserity Businesss School. He is interested in far too many things for his own good, but including social simulation, complexity, context, social intelligence, and alternative ways of distributing/organising society.
  • Nigel Gilbert is professor of Sociology at the University of Surrey and editor of the Journal of Artificial Societies and Social Simulation. He has written on the methodology of agent-based modelling and authored two textbooks on social simulation, as well as directing a number of large projects that used agent-based models.
  • Edmund Chattoe-Brown’s research addresses decisions with significant social components. Flows of information/influence through networks are obvious examples. He is interested in how agent-based modelling can systematically be informed by data routinely collected in social science, steering between data free “toy” models and “number crunching” for existing theories.

Register here: http://www.simian.ac.uk/courses/course-registration