Events and Seminars : 2013 Seminars

CURRENT PERSPECTIVES ON BIG DATA AND STATISTICAL PREDICTION

JENNIFER CLARKE, PH.D.
Associate Professor
University of Nebraska-Lincoln
TUESDAY, JANUARY 14, 2014
11:00 a.m.– 12:00 p.m.
Clinical Research Building, CRB 692

This talk comes in two parts. The first part focuses on how the world of information has changed as a consequence of the unprecedented sizes of data sets currently being collected. Two key points stand out:
1) There is a large and growing demand for people who have skills that are a mix of statistics, mathematics, computer science and various subject matter specialties such as the digital arts. Second, the mismatch between current undergrad training and the demands graduates are expected to meet– both intellectual and professional – is large, growing larger, and provides an opportunity for those willing to tear down the barriers between hitherto separate fields. The implications of this new reality are presented in the context of the University of Nebraska at Lincoln. UNL has committed resources to a Big Data Initiative to transform its research infrastructure and standard undergraduate curriculum to include ‘big data’. This includes a proposal to develop a School of Information (iSchool) at UNL.
The second part of this talk will be devoted to recent developments in predictive methods — arguably the most important statistical procedures in a big or complex data world. Even at this point, many statisticians seem unaware of the challenge of prediction, why it is important, and the interesting perspective one finds when predictive performance is used in statistical contexts that have typically relied primarily on models. This is exemplified by time series where the importance of prediction is apparent.