Machine Learning Using Tensorflow: Part Deux

Presenter: Aaron Saxton, Data Scientist, Blue Waters project, University of Illinois

Date: April 4, 2018




In the previous webinar (“Machine Learning on Blue Waters Using Tensorflow with the Image Feature Detection Problem”) we spent a lot of time discussing the “Image Feature Detection Problem” leaving ourselves almost no time to talk about Tensorflow itself. In this webinar, we will pick up where we left off. We will first use distributed Tensorflow to do a simple task of training a linear regression example. With the lessons we learn from the regression, we run a distributed Tensorflow job that uses multiple GPUs and trains an Inception v3 model on ImageNet data.

Time permitting, we will discuss possible optimizations of data ingestion techniques, batch size choices, and CPU/GPU balance.

Target Audience: Researchers interested in machine learning

Prerequisites: General scientific/high-performance computing background


Aaron Saxton is a Data Scientist who works in the Blue Waters project office at the National Center for Super Computing Applications (NCSA). His current interest is in machine learning, data, and migrating popular data/ML techniques to HPC environments. Aaron’s career has shifted back and forth between industry and academic ventures. Most recently he was a data scientist and founding member of the agricultural data company Agrible Inc. Before that, Aaron worked at Neustar Inc, University of Kentucky, and SAIC. In the summer of 2014, shortly after joining Neustar, Aaron graduated from University of Kentucky to earn his PhD in Mathematics by studying Partial Differential Equations, Operator Theory, and Schrödinger’s equation.