[PAST EVENT] Mathematics Colloquium and EXTREEMS-QED Lecture: Guannan Wang (University of Georgia)

March 20, 2015
3pm - 4pm
Jones Hall, Room 301
200 Ukrop Way
Williamsburg, VA 23185Map this location
Each day in our lives we are breathing the air of digital data. Nowadays, "Big Data" is seemingly generated at all times by everything around us. "Big Data" comes to us with great promise, as it can enhance and improve sample estimates by providing a huge number of auxiliary variables correlated with our primary variables of interest. We have entered an era where data collection is cheap, but extracting useful information from such data is not.

In this talk, a general strategy for variable selection from large data-sets is introduced based on an appealing dimension reduction tool single-index model. A non-parametric independence screening method is implemented to reduce the huge number of auxiliary variables into a moderate size. In the meanwhile, a penalized single-index model approach is proposed to simultaneously select significant variables and estimate model coeffcients. The proposed estimators are shown to enjoy the "oracle" property. A fast and efficient iterative algorithm is developed to estimate parameters and select significant variables simultaneously. The finite sample behavior of the proposed method is evaluated with simulation studies and illustrated by several real data applications.

Sarah Day