Title: "Fast subset scanning for scalable event and pattern detection"

Daniel B. Neill, Carnegie Mellon University

Abstract:

This talk will present a novel framework for accurate and computationally efficient event detection in massive, complex, and high-dimensional datasets. Our "fast subset scan" approach treats event detection as a search over subsets of data records and attributes, finding the subsets which maximize some score function (e.g., a likelihood ratio statistic). Many commonly used functions can be shown to satisfy the "linear-time subset scanning" property, enabling exact and efficient optimization over subsets. In the spatial setting, we demonstrate that proximity-constrained subset scans substantially improve the timeliness and accuracy of event detection, identifying emerging outbreaks of disease two days faster than existing methods. Fast subset scan also allows efficient integration of spatio-temporal information from a large number of data streams, both enhancing detection and improving situational awareness. Time permitting, I will also discuss our recent extensions of fast subset scan to other data types (graph/network data, tensor data, and text data), and describe our ongoing deployments of fast subset scan and other event detection methods in the public health, law enforcement, and health care domains.

Bio:

Daniel B. Neill is the Dean's Career Development Professor and Associate Professor of Information Systems at Carnegie Mellon University's Heinz College, where he directs the Event and Pattern Detection Laboratory and the Joint Ph.D. Program in Machine Learning and Policy. He holds courtesy appointments in Machine Learning and Robotics at CMU and is an adjunct professor in the University of Pittsburgh's Department of Biomedical Informatics. He received his M.Phil. from Cambridge University and his M.S. and Ph.D. in Computer Science from CMU. His research focuses on machine learning and event detection in massive datasets, with applications ranging from medicine and public health to law enforcement and security. His detection methods have been incorporated into deployed disease surveillance systems throughout the world, and his "CityScan" software is in day-to-day operational use by police to predict and prevent emerging hot spots of violent crime. Dr. Neill was the recipient of an NSF CAREER award and an NSF Graduate Research Fellowship, and was recently named one of the "top ten artificial intelligence researchers to watch" by IEEE Intelligent Systems.