RES.LL-005 | January IAP 2020 | Undergraduate

Mathematics of Big Data and Machine Learning

Instructor Insights

Valuing Concept Transition

In this section, Dr. Jeremy Kepner shares why he values the application of the concepts behind D4M by users to solve problems in their unique environments over the use of the software in its fixed form.

"I’ve come to realize that concept transition […] is perhaps more valuable and more probable than technological transition […]."
— Jeremy Kepner

D4M is the technological instantiation of the concepts that were taught in this course. People can use these concepts, just like other mathematical concepts, to write other programs or to solve problems in other domains. In fact, I’ve come to realize that concept transition (getting people to use the concepts behind the technology) is perhaps more valuable and more probable than technological transition (i.e. getting people to use the technology as-is).

Technological transition can be difficult to achieve because the technology is often in a fixed form, while the people using it are operating in unique and dynamic environments. It may be impossible to adapt your technology to meet those unique needs. But having people take the ideas behind the technology and apply them in their own environments is much easier to do. It allows users to select the ideas that work for them—they don’t need to use the technology wholesale.

When we first developed the software, we wanted everyone to download it and use it in its complete form. Many people did, and still do, that, which is wonderful. But we’ve found more people have taken the ideas behind the software and applied them to their own problems. This has worked out very well, and has really helped me see the value of concept transition.

Course Info

As Taught In
January IAP 2020
Learning Resource Types
Lecture Videos
Lecture Notes
Instructor Insights