Speaker
Shu-Wei Yeh
(National Tsing Hua University)
Description
DeepClean is a machine learning-based noise subtraction method that cleans GW strain data by subtracting noise from PEM channels monitoring environmental noises. We compiled the PEM channels to reduce AC power noises from KAGRA’s main power supplies. These channels were used to address AC power noises in KAGRA O4 strain data during offline analysis. In this study, we use DeepClean to remove AC power noise from KAGRA O4a strain data. Additionally, real-time DeepClean will be prepared for O4 as a low-latency pipeline in LVK.
Primary author
Shu-Wei Yeh
(National Tsing Hua University)
Co-authors
Chia-Jui Chou
(National Yang Ming Chiao Tung University)
Hong-Yin Chen
(National Yang Ming Chiao Tung University)
Takaaki Yokozawa
(ICRR)
Tatsuki Washimi
(NAOJ)
Albert Kong
(National Tsing Hua University)
Yi Yang
(National Yang Ming Chiao Tung University)
