Friday, October 19, at 327 Yost
Refreshments: 3:30 - 4:00 p.m, Talk: 4:00
- 5:00 p.m.
In statistical analysis and modeling we usually suppose the underlying
data are normally distributed. But in many practical cases we meet outliers
that transport important information. In some of these cases we say the
distribution of the data set obeys a heavy-tailed distribution.
Estimation of tail indices
of heavy-tailed distributions is one of the most important task of analyzing
associated data set.
In this talk, we introduce
the background of the estimation problems of tail indices, our new estimtor
of the tail-index with U-statistic structure and some other estimation
methods. As a special case, the estimation of parameters of the so-called
$\alpha$-stable distributions will be introduced.