Padhraic Smyth is recognized for his contributions to both the theory and application of probabilistic and statistical approaches to data mining.
Smyth’s research is in the area of statistical data mining and machine learning. His research focuses on both the basic principles of inference from data (theory and algorithms), combined with applications to a variety of data-driven problems in the sciences, medicine, and engineering. A central theme in his work is the use of probabilistic hidden variable models and unsupervised learning for modeling complex data.
He has made contributions in the areas of mixture models, hidden Markov models, graphical models, model-based clustering, clustering of curves and sequences, topic models for text data, pattern and event detection in time-series, and model selection. He is also well known for his work on applying these techniques to a variety of application problems involving large data sets.
For example, in climate and planetary science he has worked on clustering of storm tracks in the Atlantic and Pacific oceans, seasonal forecasting of rainfall in the tropics, analysis of global geopotential height patterns, and classification algorithms for detecting volcanoes in images of Venus. Other applications include clustering of user navigation patterns on Web sites, spatial modeling of brain images, analysis of time-course gene expression data, and event detection in large-scale traffic sensor data.
Smyth received a first class honors degree in Electronic Engineering from University College Galway (National University of Ireland) in 1984, and the MSEE and PhD degrees from the Electrical Engineering Department at the California Institute of Technology in 1985 and 1988 respectively. From 1988 to 1996, he conducted research at NASA's Jet Propulsion Laboratory. Since 1996 he has been at UC Irvine where he is currently Professor in the Department of Computer Science, with joint appointments in the Department of Statistics and in the Department of Biomedical Engineering, and is a member of the Institute for Mathematical Behavioral Sciences, the Institute for Genomics and Bioinformatics, and the Center for Research on Information Technology and Organizations. He is also the founding director for the Center for Machine Learning and Intelligent Systems at UC Irvine.
He is a coauthor of a graduate text in data mining, Principles of Data Mining, MIT Press, with David Hand and Heikki Mannila, and is also co-author of Modeling the Internet and the Web: Probabilistic Methods and Algorithms, Wiley, 2003 (with Pierre Baldi and Paolo Frasconi). He was co-editor of Advances in Knowledge Discovery and Data Mining, AAAI/MIT Press, 1996 (with Usama Fayyad, Gregory Piatetsky-Shapiro, and Samy Uthurusamy). He has served as an associate editor for the Journal of the American Statistical Association, the IEEE Transactions on Knowledge and Data Engineering, and the Machine Learning Journal, and has served as an editorial board member for the Journal of Data Mining and Knowledge Discovery and the Journal of Machine Learning Research.