Business Intelligence SIG: MAD Skills: New Analysis Practices for Big Data
MAD Skills: New Analysis Practices for Big Data
As massive data acquisition and storage becomes increasingly affordable, a wide variety of enterprises are employing statisticians to engage in sophisticated data analysis. In this talk we highlight the emerging practice of Magnetic, Agile, Deep (MAD) data analysis as a radical departure from traditional Enterprise Data Warehouses and Business Intelligence.
We present our design philosophy, techniques and experience providing MAD analytics for one of the world's largest advertising networks, using a parallel database system. We describe database design methodologies that support the agile working style of analysts in these settings. We present data parallel algorithms for sophisticated statistical techniques, with a focus on density methods. Finally, we reflect on database system features that enable agile design and flexible algorithm development using both SQL and MapReduce interfaces over a variety of storage mechanisms.
Speaker: Brian Dolan
Brian is an experienced business analyst with advanced mathematical training with a Masters in Biomathematics from the David Geffen School of Medicine at UCLA and a Masters in Pure Mathematics from the University of California, Los Angeles. Brian has more than 17 years of analytic experience, and has worked at some of most important institutions in the United States--Yahoo, MySpace, UCLA. His expertise is turning massive peta-scale data sets, from multi-terabyte databases, into strategic decisions and in mapping complicated mathematical models into actionable business solutions. His mathematical specialties are in time series analysis, longitudinal analysis, applied stochastic processes and optimization. He has also provided statistical and modeling expertise for projects in the public interest applying statistical, solution-driven methodology to deal with election fraud issues, falcon ecology and hospital admission forecasting. He is a frequent speaker on Machine learning topics at Institute Business Forecasters, UCLA, UC Berkeley, and SAS and is the co-author of the Mad Skills paper.
6:30pm - 7:00pm - Registration
7:00pm - 8:30pm - Presentation