Hankel low-rank matrix approximation for time series denoising in gravitational-wave science

ORAL

Abstract

Extracting overlapping periodic signals from noisy time-series data is a common challenge in gravitational-wave astronomy and pulsar timing. We present an approach that recasts this problem into one of low-rank matrix approximation by embedding the time series into a Hankel matrix. We demonstrate the effectiveness of this method through numerical experiments on synthetic datasets using two iterative algorithms: Cadzow and IRLS.

*This research was partially supported by a seed grant from the Space@Hopkins initiative at Johns Hopkins University, by NSF Grants No. AST-2307146, No. PHY-2513337, No. PHY-090003, and No. PHY-20043, by NASA Grant No. 21-ATP21-0010, by John Templeton Foundation Grant No. 62840, by the Simons Foundation [MPS-SIP-00001698, E.B.], by the Simons Foundation International [SFI-MPS-BH-00012593-02], and by Italian Ministry of Foreign Affairs and International Cooperation Grant No. PGR01167. The authors acknowledge the computational resources provided by the WVU Research Computing Thorny Flat HPC cluster, partly funded by NSF OAC-1726534.

–

Presenters

  • Vladimir Strokov

    • West Virginia University

Authors

  • Nicholas Geissler

    • New York University

  • Vladimir Strokov

    • West Virginia University

  • Sergey Kushnarev

    • Johns Hopkins University

  • Christian Kuemmerle

    • University of Central Florida

  • Emanuele Berti

    • Johns Hopkins University