Improving gravitational wave search sensitivity with TIER: Trigger Inference using Extended strain Representation

2025-07-11

AI Summary

Near-threshold candidate events in gravitational-wave searches suffer from insufficient signal-to-noise ratio (SNR), hindering reliable identification of high-fidelity and unequal-mass binary systems. Method: We propose TIER, a machine learning framework that leverages sparse temporal features extracted from ~10-second extended strain data surrounding each candidate—without relying on simulated signal injections—to train a lightweight classifier directly mapping candidates to their probability of being astrophysical gravitational-wave events. Contribution/Results: TIER seamlessly integrates its output into existing search pipelines’ significance estimation, substantially enhancing the detection statistic. Applied to LIGO–Virgo–KAGRA O3 data, it achieves up to a 20% gain in sensitive volume-time, particularly improving detection sensitivity for mergers in the mass gap and intermediate-mass black hole binaries. This provides a more robust candidate filtering tool for multi-messenger astronomy.

Technology Category

• Machine Learning: Hardware-aware ML
• Search and Optimization: Algorithm Configuration
• Humans and AI: Human-in-the-loop Machine Learning

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Web Mining and Content Analysis: Machine learning and data science for the Web
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

We introduce a machine learning (ML) framework called $ exttt{TIER}$ for improving the sensitivity of gravitational wave search pipelines. Typically, search pipelines only use a small region of strain data in the vicinity of a candidate signal to construct the detection statistic. However, extended strain data ($sim 10$ s) in the candidate's vicinity can also carry valuable complementary information. We show that this information can be efficiently captured by ML classifier models trained on sparse summary representation/features of the extended data. Our framework is easy to train and can be used with already existing candidates from any search pipeline, and without requiring expensive injection campaigns. Furthermore, the output of our model can be easily integrated into the detection statistic of a search pipeline. Using $ exttt{TIER}$ on triggers from the $ exttt{IAS-HM}$ pipeline, we find up to $sim 20%$ improvement in sensitive volume time in LIGO-Virgo-Kagra O3 data, with improvements concentrated in regions of high masses and unequal mass ratios. Applying our framework increases the significance of several near-threshold gravitational-wave candidates, especially in the pair-instability mass gap and intermediate-mass black hole (IMBH) ranges.

Problem

• Improving gravitational wave search sensitivity using extended strain data
• Enhancing detection with ML on sparse extended data features
• Boosting significance of near-threshold gravitational-wave candidates

Innovation

• ML framework using extended strain data
• Sparse summary features for efficient training
• Improves detection sensitivity without injections