Time Series Center   Harvard University

Pavlos Protopapas received his PhD in 1996 at the University of Pennsylvania in theoretical nuclear physics. His thesis provided a solution to the Coriolis attenuation problem. He served as the associate director of the National Scalable Cluster Project (NSCP), one of the initial attempts at large scale distributing computing on a grid-like model. Protopapas is a member of the outer solar system team for Pan-STARRS and the TAOS project. His research interests are in planetary transits, the outer solar system, photometric variability, microlensing and in computer science on large databases and data mining in astronomy in particular in feature extraction, anomaly detection, and similarity searches in time series. more »

Charles Alcock is the Directer of the Harvard-Smithsonian Center for Astrophysics (CfA) in Cambridge, MA. Through his initiation and leadership of the MACHO Project, took astronomical analysis into the multi-terabyte regime. This project is well known for the discovery of gravitational microlensing, the rare magnification of the light of distant stars by otherwise invisible massive objects. At the end the MACHO Project, Alcock initiated the Southern Edgeworth-Kuiper Survey, using the same hardware to search for new solar system objects. He is the US PI for the TAOS Project, which will be one of the test-beds for the time series center. more »

Roni Khardon obtained his Ph.D. in Computer Science from Harvard University, under Prof. Leslie Valiant. He joined the Computer Science department at Tufts University in Fall 2000, where he is now an Associate Professor. Prior to moving to Tufts in 2000 he held a faculty position in the University of Edinburgh in Scotland. Khardon's interests include theoretical foundations, efficient algorithms, practical aspects and applications of machine learning, data mining and knowledge representation. Concrete problems range from learning to classify objects, learning in order to act, and data mining of frequent patterns. A particular focus over the last few years has been solving these in contexts where objects and relations give a natural way to describe the problems. more »

Carla E. Brodley is a professor in the Department of Computer Science at Tufts University. She is currently serving as acting chair of the Department of Computer Science at Tufts. She received her PhD in computer science from the University of Massachusetts, at Amherst in 1994. From 1994-2004, she was on the faculty of the School of Electrical Engineering at Purdue University. She researches fundamental machine learning and data mining issues, including anomaly detection, feature selection for unsupervised learning, clustering of high dimensional data and correlation analysis. In particular, she is well known for the successful application of machine learning and data mining to the areas of remote sensing, content-based image retrieval, covert-channel detection in networks, and user behavioral modeling for intrusion detection. more »

Eamonn Keogh’s research areas include machine learning and information retrieval, specializing in techniques for solving similarity and indexing problems in time-series datasets. He has authored more than 90 papers. He received the SIGMOD 2001 best paper award, and runner up best paper award in KDD 1997. In 2000, he was nominated as a teaching assistant consultant at UC Irvine, where he received a campus-wide award for teaching excellence in 1998. While completing his Ph.D. degree, he worked as a head teaching assistant, lecturer and research assistant. Dr. Keogh is a member of ACM, SIGMOD, SIGKDD and AAAI. more »

Rahul Dave is a computational scientist at the IIC. He receiving his doctorate in Astrophysics from the University of Pennsylvania in 2002. His thesis included the first systematic study of the effects of dark energy on the expansion history and fluctuation development in the universe. He was Associate Director of the Eniac2000 project, an early large scale commodity cluster effort at the University of Pennsylvania. He is a member of the TAOS project, which is looking for TNOs in the outer solar system. His current research efforts are directed to the development of search and database technology to make almost real-time queries on metadata and morphology aspects of large databases of time-series possible.

Patrick Ohiomoba is a Sr. Systems Administrator at the IIC. He holds a B.A. in Applied Mathematics (with a specialization in Chemistry) from Harvard University, and he is currently a candidate for the S.M. degree in the School of Engineering and Applied Sciences at Harvard. Patrick's professional interests include open source technologies, scientific computing, agile web applications frameworks, probabilistic informatics, distributed/parallel computing, and systems programming/engineering.

Gabriel Wachman is a Ph.D. candidate at Tufts University, working under Prof. Roni Khardon. His research focuses on learning from structured data, and in understanding the effects of margin on learning algorithms. more »

Steve Pember is a Masters/PhD student at Tufts University, working with Carla Brodley. His current research focuses on improving techniques for fast similarity search of light curves.

Umaa is currently a Ph.D. student in the computer science department at Tufts University. Her advisor is Carla Brodley. She received a B.A. in Mathematics from the University of California, Berkeley and worked as a software engineer at the San Francisco-based internet media company CNET Networks. For her master's research, she developed a technique to identify mislabeled instances in training data by assigning a confidence on the instance's cleanliness and using that confidence as a weight during classifier training. Her area of research is machine learning/data mining. She currently focuses on the problem of anomaly detection in photometric time series. more »

Dae-Won is a visiting student from Yonsei student in Seoul, Korea, where he is pursuing his Ph.D. under Professor Yong-Ik Byun. He is a member of the TAOS collaboration. He will be working with Dr. Pavlos Protopapas and Prof. Charles Alcock at the Smithsonian Astrophysical Observatory(SAO) and the IIC on planetary transits.

Taryn Nihei is a student of Charles Alcock at the University of Pennsylvania and the CfA in Cambridge, MA. She is presently writing her doctoral thesis on light curve analysis.

Dan is a recent Brandeis University graduate, majoring in Computer Science with a senior thesis study involving Time Series. His research involves time series analysis, currently using methods in scan statistics and optimization algorithms.

Federica Bianco is Smithsonian predoctoral fellow a graduate student at the University of Pennsylvania. Her work involves searching for short rare events in a high cadence time series. In particular searching for trans-neptunian bodies occultation signatures with the goal of producing constraints on solar system formation models. more »

Samuel Chen was a summer Intern at the TSC at IIC in the summer of 2007. He worked on wavelength techniques in the analysis of time series. He is currently teaching in China.

Evan Morikawa is a sophomore at Olin College in Needham, MA. He was an intern at the TSC at IIC in the summer of 2007. He developed Filterflow, an innovative workflow application for applying filters to light curves in a reusable, example-to-batch fashion, fronted by a simple and intuitive GUI. more »

Davis Kosslyn is a freshman at Harvard University. He was an intern at the TSC at IIC in the summer of 2007. He worked on database enabling the TAOS light curves.

John Rice is Professor of Statistics at the University of California at Berkeley. He is interested in formulating methods for analyzing data that arise in the form of random functions, such as time series, and which involve large quantities of data and computationally intensive analysis. His work with graduate student James Reimann led to algorithms for identifying variable stars in the MACHO data base. Much of his recent work has centered around two projects in astronomy: detecting objects in the outer regions of the solar system (the Kuiper Belt) and detecting gamma-ray pulsars. The probability of occultation events is so low that it will be necessary to conduct 100 billion measurements per year in order to detect the ten to four thousand such occultation events expected. Thus, foremost among the statistical problems is the necessity of developing methods to detect very rare, faint events in very large quantities of data. The statistical challenge in gamma-ray pulsars to infer from a sequence of arrival times of photons, whether the source is periodic, corresponding to a pulsar, or whether it is constant, corresponding to background radiation. more »

Rosanne Di Stefano is an astrophysicist at the CfA. Her research interests are in microlensing and mesolensing, the study of close binary systems, and X-ray emission from galaxies. She was trained in high-energy theoretical physics. She became interested in time series analysis during the early 1990s, when she began to participate in studies of non-linear dynamics at a workshop at the Aspen Center for Physics. Her first project in astrophysics was to apply nearest-neighbor phase space techniques to the study of astronomical data sets. Although work on certain well-studied variable stars had demonstrated feasibility, most efforts by the small community of people working on such projects met with only limited success, because the quality of the data was not yet good enough. The well-sampled data sets we will work with, will allow these types of investigations to be successful.