Seminar: Scalable Machine Learning for R, by Joseph Blue, Director Global Data Science, MapR

posted Dec 10, 2016, 11:36 AM by Yanchang Zhao   [ updated Dec 10, 2016, 11:38 AM ]

Title: Scalable Machine Learning for R with MapR
Speaker: Joseph Blue, Director of Global Data Science for MapR
Time and Date: 4:30-6:30pm, Thursday, 15 December 2016
Location: 6C35 (building 6, room C35), University of Canberra

In this discussion, we will review some of the popular libraries that allow R users to interact with large-scale and unstructured data sets through the use of a distributed environment. After a brief intro to deep learning concepts, a demo will be shown that leverages H2O from R to detect anomalies in recent Australian stock prices. The RMarkdown notebook and other resources will be provided to attendees. Additionally, any questions from the audience involving the use of R in distributed environments can be addressed.

Joe is the Director of Global Data Science for MapR, where he has focused on collaborating with customers to explore large, unstructured data sources and derive real business value for the past 3 years. He has deployed solutions in financial, healthcare, advertising, manufacturing, retail and media markets.

Prior to joining MapR, Joe developed predictive models in health care for Optum (a division of UnitedHealth) as Chief Scientist. He was the first Fellow for Optum's start-up, Optum Labs and has several patents pending.

Before his time at Optum, Joe accumulated over 10 years of data science experience at Fair Isaac, Lexis Nexis, HNC Software and ID Analytics (now LifeLock) specializing in business problems such as fraud & anomaly detection, which yielded a patent for identity theft detection.