Normalizing Flow Emulator for Gravitational Wave Detection Probability in the LIGO/Virgo/KAGRA O4 Observing Run

Abstract

Efficient and accurate estimation of the detection probability of a gravitational wave event given its source parameters remains a challenging problem because it conventionally requires expensive Monte Carlo injection campaigns. We present a machine learning-based approach using Normalizing Flows (NFs) to estimate the detection probability of gravitational wave events for a LIGO/Virgo/KAGRA O4-like detector network configuration. We utilize the publicly available injection dataset as training an…