Abstract:
The last five years have seen an explosion in the number of studies using data to identify interactions within groups. Approaches from machinelearning, Bayesian inference and statistical mechanics have been employed to great effect to learn how individuals respond to each other and their environment. As data gets ever "Bigger", and we seek to understand more subtle and complex effects, robust statistical inference becomes ever more important. In this talk I will reflect on some of the key statistical problems that have emerged in this effort to understand collectives, and potential solutions for future studies. Topics include: the ecological fallacy, emergence, multiple hypothesis testing and time series. |