Distance-based variable generation with applications to the FACT experiment. Issue 7 (18th May 2016)

• Record Type: Journal Article
• Authors: Voigt, Tobias; Fried, Roland; Rhode, Wolfgang; Temme, Fabian
• Is Part Of: Journal of applied statistics. Volume 43:Issue 7(2016)
• Page Start: 1186
• Page End: 1197
• Publication Date: 2016-05-18
• Subjects: classification -- astrophysics -- Hillas -- random forest -- FACT; Statistics -- Periodicals; 519.5
• DOI: 10.1080/02664763.2015.1092115
• Languages: English
• ISSNs: 0266-4763

Abstract

Abstract : We introduce a new way to construct variables for classification in a setting of astronomy. The newly constructed variables complement the currently used Hillas parameters and are specifically designed to improve the classification. They are based on fitting elliptic or skewed bivariate distributions to images gathered by imaging atmospheric Cherenkov telescopes and evaluating the distance between the observed and the fitted distribution. As distance measures we use the Chi-square distance, the Kullback–Leibler divergence and the Hellinger distance. The new variables lead to an improved classification in terms of misclassification errors.