STSM Evaluation of statistical properties of large data sets

Workplan: We plan to examine the problem of how to evaluate variety of statistical quantities and functions from very large data sets linked to economics and finance. We propose using financial data as the exemplar but the methods will also be explored in the context of other socio-economic data sets (for example multiple choice examination results) where statistical properties play an important role and are of interest.
In Bulgaria Oleg Yordanov presently has one student looking at simulation of random processes with predefined distribution and correlation properties. We intend to explore simulations of nonlinear time series and nonlinear spatial random processes (fields). We have codes for random processes with power-law frequency spectra, known to have nontrivial fractal properties. Simulation of such data can be used to ascertain variety of models, including, models in information and social sciences. The aim is to discuss and develop these matters during the short term mission. We shall also visit Plovdiv U who are developing a MS program in social sciences, much in a fashion of Econophysics and/or Sociophysics. Peter has been invited to talk on these issues using illustrations from his experience and recent textbook on the matter.