Leland Wilkinson Talk - High-Dimensional Visual Analytics: Exploring structure using low-dimensional projections

Date
–

SPEAKER: Leland Wilkinson, Computer Science, UIC / Skytree Inc.

HOST: Tamara Munzner

LOCATION: ICICS/CS X836, 2366 Main Mall, Department of Computer Science (Map) 

ABSTRACT

This talk covers a research program that rests on algorithms for computing geometric graphs on low-dimensional projections. These algorithms have led to:

  1. discovering canonical structures and anomalies in collections of 2D scatterplots
  2. discovering anomalies in high-dimensional point sets
  3. a random-projection ensemble classifier that outperforms decision trees, random forests, and support vector machines on widely-used test datasets
  4. a system for analyzing visually massive multivariate time series
  5. a system for clustering images
  6. an expert system for statistical analytics that discovers and remedies statistical irregularities.

Software from these projects will be presented.

BIO

Leland Wilkinson is VP of Analytic Visualization at Skytree Inc. (https://skytree.net/wp-content/uploads/2018/11/team.jpg) and Adjunct Professor of Computer Science at the University of Illinois at Chicago (https://www.cs.uic.edu/~wilkinson/). He wrote the SYSTAT statistical package and founded SYSTAT Inc. in 1984. Wilkinson is a Fellow of the American Statistical Association, a Fellow of the American Association for the Advancement of Science, and an elected member of the International Statistical Institute. In addition to journal articles and the original SYSTAT computer program and manuals, Wilkinson is the author of The Grammar of Graphics, the foundation for a number of commercial and open-source visual analytic systems (IBM-RAVE, Tableau, R-ggplot2, and Python-Bokeh).

Tags
research