XuanLong Nguyen Assistant Professor, Department of Statistics, University of Michigan ("Long Nguyen" by friends and colleagues) Long's Email: xuanlong@umich.edu Office: 445B West Hall Phone: 734-763-3499, Fax: 734-763-4676 Mail Address: 439 West Hall, 1085 South University, Ann Arbor, MI 48109-1107

[Publications] [Research] [Teaching]

Research interests: statistical and computational modeling, machine learning, optimization methods for structured data and distributed systems

• Distributed statistical inference, variational inference methods
• Nonparametric Bayesian modeling for functional and spatial data
• Image and signal processing, text processing
• Detection, tracking and estimation applications in distributed systems, sensor networks
• Applications of modeling and computation in the environmental science

Research synopsis

Twenty-first century technology and science confront statisticians and engineers with data of unprecedented scale and heterogeneity. Moreover, data are often collected using large distributed networks of diverse monitoring devices. In broad terms, my research interests lie in addressing these issues of scale, heterogeneity, and distributed monitoring, particularly as they arise in a range of interdisciplinary application areas, including signal processing applications in distributed sensor networks, as well as ecological modeling and computation in the environmental sciences.

From a methodological standpoint, I am interested in the development of models based on nonparametric, spatial and graphical statistics, with a particular concern for the issues that arise when we consider the distributed nature of the underlying data collection and data processing systems. I am also interested in the development of new deterministic and sampling-based inferential methods for complex models, particularly those motivated from the viewpoint of convex analysis and constrained optimization. Examples include variational inference for graphical models, flexible estimation methods based on convex loss functions for distributed statistical inference, and surrogate posterior distributions for nonparametric Bayesian models.

Some recent papers: Go to List of publications

• The Dirichlet labeling process for functional data analysis. X. Nguyen and A. E. Gelfand. Technical Report 08-37, Dept. of Statistical Science, Duke University, November 2008. Submitted.
• On surrogate loss functions and f-divergences. X. Nguyen, M.J. Wainwright and M.I. Jordan. Annals of Statistics, 37(2), 876--904, 2009.
• Estimating divergence functionals and the likelihood ratio by convex risk minimization X. Nguyen, M.J. Wainwright and M.I. Jordan. Technical Report 764, Dept. of Statistics, UC Berkeley, Sept 2008. [ arxiv link ]. Submitted.
• On optimal quantization rules in some problems in sequential decentralized detection. X. Nguyen, M. J. Wainwright and M. I. Jordan. IEEE Transactions on Information Theory, 54(7), 3285--3295, 2008.
• Nonparametric decentralized detection using kernel methods. X. Nguyen, M. J. Wainwright and M. I. Jordan. IEEE Transactions on Signal Processing, 53(11), 4053--4066, 2005.
• A kernel-based learning approach to ad hoc sensor network localization. X. Nguyen, M. I. Jordan and B. Sinopoli. ACM Transactions on Sensor Networks, 1, 134--152, 2005.