K-Means Clustering: Algorithm and Its Properties

Consider the problem of finding close-knit communities in social networks. Or the problem of using computer vision to aid self-driving cars. Or even the problem of recommending movies to the customers of an online video streaming service. One technique that can be useful in tackling all these seemingly diverse problems is the unsupervised learning technique called clustering. Clustering is a machine learning method that groups points into clusters based on some notion of distance. In this talk, I will focus on the k-means clustering algorithm. I will provide both an intuitive as well as a more formal look at the algorithm and some of its properties. I will also highlight some connections of this algorithm to my research in prediction markets.
Sindhu Kutty is an Assistant Professor in the Department of Mathematics and Computer Science at the University of Detroit Mercy. Prior to this, she was a Visiting Assistant Professor in the Department of Computer Science at Swarthmore College. She obtained her Ph.D. from the University of Michigan in 2015 where she has assisted in teaching a variety of courses. Her research interests lie in the design and analysis of social computing systems, with a focus on market mechanism design and its connections to statistical machine learning. She is passionate about teaching and mentoring undergraduate student research.