The Sci-Files – 01/26/2020 – Tyler Derr – Graphs and Deep Learning

Chelsie Boodoo and Daniel Puentes

Hang on for a minute...we're trying to find some more stories you might like.


Email This Story






On this week’s The Sci-Files, your hosts Chelsie and Danny interview Tyler Derr.

Tyler is a 5th year Ph.D. student in the Computer Science and Engineering and a member of the Data Science and Engineering Lab. His research is generally focused on data science where he seeks to extract insightful patterns in data that can then be used to both understand/analyze the past and to make predictions about the future. More specifically, due to the fact that much of today’s big data can be represented as graphs, his emphasis has primarily been on harnessing this natural structure of data. In this case, when we mention graphs, we are talking about a collection of objects that have relations between them and can furthermore be associated with a set of features/attributes associated with both the objects and their relations.

An example of a graph can be a social network consisting of online users as the objects, different user relations such as “friends”, “blocked users”, etc., and attributes such as their “likes”. His research is currently focused on how to advance state of the art deep learning approaches (which is an area of machine learning that utilizes artificial neural networks) to harness the graph structure of data. For example, if seeking to categorize the political affiliation of an online user, can we come up with a way of incorporating those you’ve followed (and unfollowed) in a deep learning framework to obtain better predictions.

If you’re interested in talking about your MSU research on the radio or nominating a student, please email Chelsie and Danny at scifiles@impact89fm.org. Check The Sci-Files out on Twitter @SciFiles89FM and Facebook!