Multiscale stochastic systems in the social sciences

Andrea Bertozzi
University of California, Los Angeles (UCLA)

I will review several classes of models from the recent literature related to routine activity of individuals associated with specific actions. Some examples involve crime modeling in which comparison with data from local law enforcement can be done. More recently predictive modeling has led to commercial software packages now in use across the US in an a few foreign cities. I will also discuss stochastic dynamics in electronic media such as email traffic and online social media such as twitter. Analysis of such data is very new and there are opportunities for scientists to provide new ideas and insights into this growing area of research.
Sotchastic models include basic biased random walk models for spatial movement and self-exciting point process models for decision making activity. We will examine different classes of activities and how one might classify the statistical pattern of behavior.

Back to Uncertainty Quantification for Multiscale Stochastic Systems and Applications