This project seeks to provide automated methods for analyzing sky surveys to detect extra-solar planets via anomaly and novelty detection methods. Extremely large astronomical synoptic surveys will soon monitor much of the sky regularly, detecting vast numbers of interesting, variable astronomical objects. The objective of this proposal is to develop the tools necessary to exploit these new data in order advance discovery. The proposed work will explore the scientific analysis and modeling of massive datasets of light-curves (66 million light-curves now available, growing to 100 billion in a decade) from a broad range of perspectives. The unique complications posed by the data in this domain will drive research for data mining techniques to allow these analyses. A comprehensive framework of models and their relationships with the data will be developed that readily separates two kinds of interesting objects: (1) Rare objects that reveal special insights about the models; and (2) Potential truly novel objects that cannot be described by the models. New computationally tractable time series algorithms for novelty detection will be developed for these purposes.
