With 2 Speakers:

Predictive Analytics with Hadoop
Presented by Robert Chu

"Essentially, all models are wrong, but some are useful." - George E. P. Box

Predictive modeling is an iterative process of defining a model and then evaluating its usefulness. This process can easily become drawn out and cumbersome when building models with big data sets. KijiExpress is designed to make predictive modeling easier by providing much needed tooling and allowing users to define MapReduce steps through Scalding jobs. Using Enron's email data set as a use case, this talk will demonstrate how to define, train and validate predictive models using KijiExpress.

Bio:

Robert is a member of the engineering team at WibiData. He develops tools that enable data scientists and engineers to seamlessly develop and deploy real-time predictive models using machine learning and natural language processing. He graduated with a BS in Computer Engineering from the University of Washington.