Invited Seminar Talk
Prof. Jessica Lin, George Mason University (USA)
when: 14 March 2016, 3pm
where: room Turing Reunion, 7th floor, Paris Descartes University, 45 Rue Des Saints Peres, Paris 75006
Massive amounts of data are generated daily at a rapid rate. As a result, the world is faced with unprecedented challenges and opportunities on managing the ever-growing data, and much of the world's supply of data is in the form of time series. Time series data mining has thus attracted an enormous amount of attention from researchers and practitioners in the past two decades. This talk will focus on the discovery of novel and non-trivial patterns in time series data, including frequently encountered patterns (motifs) and rare patterns (anomalies). The ability to efficiently detect frequent and anomalous patterns in time series allows for the exploration, summarization, and compression of data. In addition, such information is crucial to a variety of application domains where these patterns convey critical and actionable information. In recent work, we demonstrate that grammar induction, the process of learning rules of a formal language from a set of observations, allows the discovery of hierarchical structures and regularities from input time series. We propose several algorithms based on grammar for efficient discovery of co-existing variable-length approximate motifs and anomalies without any prior knowledge about their length, shape, or minimal occurrence frequency. Based on time series motifs, we further propose an algorithm to identify class-specific representative patterns for time series classification. The key motivation is that the identification of a small set of distinctive and interpretable patterns of each class allows us to exploit their key characteristics for discriminating against other classes. We present GrammarViz, an interactive tool for grammar-driven mining and visualization of variable-length time series patterns.
Dr. Jessica Lin is an Associate Professor in the Department of Computer Science at George Mason University. She received her PhD from University of California, Riverside in 2005. Her research interests focus on the mining of large datasets, in particular, time series and spatiotemporal data. In recent years, she has worked on several collaborative projects--all of which involve the discovery of novel patterns from massive time series or spatiotemporal databases--in diverse disciplines including semiconductor manufacturing, medicine, national security, geoinformatics, and earth sciences, with support from the National Science Foundation (NSF), U.S. Army, Naval Research Lab, and Intel Corporation. Dr. Lin has been in the program committee for many international conferences in the area of data mining. She is the co-Editor of a newly published book, "Spatio-Temporal Data Mining for Geoinformatics: Methods and Applications." Dr. Lin teaches advanced topics on data mining at GMU, concentrating on the mining of multimedia data including text, image, time series, and social media.
Hosted by: Themis Palpanas
List of past seminars