Department of Mathematical Sciences
|DATE:||Thursday, October 15, 2015|
|TIME:||1:15pm to 2:15pm|
|SPEAKER:||Shih-Kang Chao, Department of Statistics, Purdue University|
|TITLE:||Quantile Regression for Extraordinarily Large Data|
One complexity of massive data comes from the accumulating errors that are often unknown and may even have varying shapes as data grows. In this talk, we consider a general quantile-based modelling that even allows the unknown error distribution to be arbitrarily different across all sub-populations. A delicate analysis on the computational-and-statistical tradeoff is further carried out based on nonparametric sieve estimation.