University of Minnesota
Software Engineering Center

You are here

Scott Ernst, Ph.D.

Photo of Scott Ernst, Ph.D.
Scott is currently the Director of Data Science & Data Engineering at When I Work, a Minneapolis-based startup. He has a PhD in computational physics that focused on large-scale astrophysical and magnetohydrodynamic plasma simulations. Over the last decade Scott has worked in various data science, visualization, engineering and architecture roles. These include leading the data science team for an international research project modeling dinosaur behavior on the world’s largest dinosaur track-site and creating 3D digital visualizations for clients all over the world such as National Geographic, National Public Radio (NPR), Carnegie Natural History Museum, Los Angeles Natural History Museum, Asahi Shimbun (朝日新聞), Tokyo Natural History Museum and Jurassica in Switzerland. Scott is also the creator of Cauldron, a data science notebook that expands notebook-based computing beyond interactive usage for production applications.

Recent Presentations

Essentials of Effective Machine Learning

We hear about machine learning successes all the time but not nearly as much about machine learning failures. Many assume this means that failures are uncommon, but there’s plenty of less publicized evidence to the contrary. As hype gives way to reality and data science matures as a discipline, it’s past time for increased scrutiny over how machine learning is carried out effectively.