Novelty Detection in Search of Anomalous Objects within the WISE All Sky Survey

摘要

Surveying the previously unexplored areas of the sky guarantees delivering information about unexpected sources whose existence or properties cannot be anticipated. Novelty detection algorithms allow to efficiently search for such objects. In this work we present the ability of one-class support vector machines (OCSVM) algorithm in service of novel source extraction within the infrared AllWISE catalog covering the whole sky.