Abstract:

A real world flux of data is usually modeled as a sequence of independent and identically distributed (i.i.d.) random variables, or by a linear (nonlinear) regression model driven by an i.i.d. sequence. However, this modeling is far from true in many real world situations. Uncertainty essentially hidden inside the real world data sequence should be taken into consideration. In this talk we introduce a robust nonlinear expectation and a simple algorithm of phi-max-mean to quantitatively measure and calculate this type of uncertainty. It is proved that the phi-max-mean is an asymptotically optimal unbiased estimator to the corresponding nonlinear distribution.

Poster (427KB)

Registration

Light refreshments will be served at 16:30 at Room 121 Lady Shaw Building

<<  All are welcome  >>