The identification of electromagnetic emission from gravitational-wave sources typically requires multiple follow-up observations due to the limited fields of view of follow-up observatories compared to the poorly localized direction of gravitational waves. Gravitational-wave localization regions are typically covered with multiple telescope pointings using a “honeycomb” structure, which is optimal only on an infinite, flat surface. Here we present a machine-learning algorithm which uses genetic algorithms along with Broyden-Fletcher-Goldfarb-Shanno optimization to find an optimal configuration of tiles to cover the gravitational-wave sky localization area on a spherical surface.

• Received 10 March 2020
• Accepted 26 May 2020

DOI:https://doi.org/10.1103/PhysRevD.101.123008