The Central Limit Theorem, Stein's Method and Spatial Random Graphs
Dr. Tara Trauthwein, University of Oxford
Abstract: The classical Central Limit Theorem states that sums of independent random variables behave more and more like normally distributed random variable when we increase the number of summands. This simple result has astonishingly wide consequences, since it means that a huge number of real phenomena observed in large quantities can be safely estimated by using the Gaussian distribution. Stein's Method, introduced in 1972, is a hugely convenient tool for the estimation of the theoretical distance between any random variable, and a Gaussian. In this talk, we will give a gentle introduction to Stein's Method and show how it can be used, not only in a context of sums of independent random variables, but in the rich topic of (spatial) random graphs.
Tea and refreshments available at 3:00 p.m. in the Assmus Conference Room (CU 212).