Speaker:
Joseph Barr, Ph.D.
Chief Scientist
ID Analytics
www.idanalytics.com
 
Abstract
With an increasing profusion of data new (statistical) modeling techniques have 
emerge. Although classical statistical techniques like multiple linear 
regression are valid and still very popular, they fall short to providing 
predictive capabilities necessary in an increasingly complex and unstructured 
universe of distinct domains.  During the past quarter of a century, new breeds 
of modeling techniques gave rise to modeling performance previously not 
imagined.  Although the “curse of dimensionality” still posing a daunting 
challenges to modelers, science & technology has developed the tools to deal 
with such unwieldy complexity.  I will described in some detail how we moved 
from regression and its variants, through LOGIT and the generalized linear 
models family to machine learning and heuristics. I will explain what 
educational tools a graduate must possess to enter into this exciting, 
challenging and rewarding workforce – to be a bona-fide “analytics 
professional”. Finally I will give a high-level account of ID Analytics, what we 
do, the kind of analytics/IT professionals we seek to hire.
 
Intended audience: (advanced) students majoring in CS, math, stat, EE. 
 
Contact: jbarr@idanalytics.com