Abstract |
This talk presents perspectives on the analysis of massive data sets. These derive from past efforts to develop tools for analysis of large data sets and speculations about the future. The perspectives touch on many topics; previous research, recurrent problems, standard tricks for addressing problems, new problems, and new opportunities. In looking forward, it seems that evolving notions about man-machine interface, models, and model criticism will play an important role in massive data set analysis. The extent to which the statistical community responds this evolution will have far reaching consequences in terms of the influence of our profession with the scientific community. |