http://arxiv.org/abs/1704.06413
The machine learning techniques are widely applied in many modern optical sky surveys, i.e. Pan-STARRS1, PTF/iPTF and Subaru/Hyper Suprime-Cam survey, to reduce the human intervention for data verification. In this study, we have established a machine learning based real-bogus system to reject the false detections in the HSC-SSP source catalog. Therefore the HSC-SSP moving object detection pipeline can operate more effectively due to the much less false positives inputs. To train the real-bogus system, we use the stationary sources as the real training set and the `flagged’ data as the bogus set. The training set contains 49 features, which, in majority, are the photometry measurements and shape moments generating from the HSC image reduction pipeline (hscPipe). Our system can reach a true positive rate (tpr) ~ 96% with a false positive rate (fpr) ~ 1% or tpr ~ 99% at fpr ~ 5%. Therefore we conclude that the stationary sources are decent real training samples, and using photometry measurements and shape moments can reject the false positives effectively.
H. Lin, Y. Chen, J. Wang, et. al.
Mon, 24 Apr 17
3/54
Comments: 16 pages, 4 figures, submitted to PASJ HSC special issue