Lead investigators
Pavlos Protopapas (Harvard Faculty of Arts and Sciences
Harvard-Smithsonian Center for Astrophysics)
Description
New astronomical surveys such as Pan-STARRS and LSST are under development and will collect petabytes of data. This data will include light curves from vast numbers of moving, variably bright, and transient objects. Such time series observations are also common to other disciplines: e.g. prices of shares and seismic observations at times of earthquakes and aftershocks. A common set of problems in the analysis of these time series concerns the issues of feature extraction, similarities, and the search for anomalies or outliers. The goal of the Time Series Center is to create a huge database of cross-referenced light curves and other time-series, and to bring together scientists from a diverse set of fields, starting with Astronomy, Computer Science, and Statistics, to research interesting algorithms to address these problems.
Our current projects are:
- Web Services based queries and workflow to access light curves from multiple surveys
- Outlier detection in large data sets of time series
- Discovery of extra-solar planets
- Discovery of Trans-Neptunian Objects
- Automatic classification of variable stars
- Light Curve Morphological Querying
- Multiple projection based database indexes
- Measuring Temporal Symmetry in Quasar Light Curves
- Variability estimation in time series through scan statistics
More information is available.
