I received my Ph.D in statistics from Stanford University in 2003, and joined the Statistics Department at Carnegie Mellon University in 2007. I was trained in statistical inference for Big Data, and specialized in dealing with the most challenging regime where the signals are so rare and weak that many conventional approaches fail and it is desirable to find new methods and theory that are appropriate for such a situation. I received the NSF CAREER award in 2007, IMS Tweedie Award in 2009, and have been an elected IMS Fellow since 2011. I delivered an IMS Medallion Lecture (2015), the IMS Tweedie Lecture (2009) and some plenary or keynote lectures.
My primary research interest is on analyzing big data where the signals we are searching for, although quite numerous overall, are sparse compared to the number of would-be false signals and are also individually weak. In this "rare and weak" setting, classical methods and most modern empirical methods are simply overwhelmed. I have been developing methods appropriate for such settings in the past years, studying problems in various areas including large-scale testing, classification, clustering, variable selection, and most recently, network analysis and low-rank matrix recovery. I am interested in statistical machine learning, social networks, genomics and genetics, and neuroscience.