Scholarlink

The field of Gravitational Wave research has exploded in recent years. Powerful laser interferometers like Advanced Laser Interferometer Gravitational Wave Observatory (LIGO) and Advanced Virgo are now listening for the universe’s most violent events. While these technological marvels have significantly improved the precision of gravitational wave data collection, the data itself is not perfect. Noise can still creep in, potentially leading to misinterpretations. One problematic type of noise in Gravitational Wave data is the glitch. We define a glitch as a noise event that can either masquerade as a real signal from space or degrade the overall quality of the data. In this chapter, we present our current work for the task of detecting Gravitational Wave glitches using Machine Learning and Deep Learning models. We also establish a clear benchmark to compare their performance, highlighting the models that achieve the best results based on three key metrics: accuracy, precision, and recall.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Efficient ML Algorithms for Detecting Glitches and Data Patterns in LIGO Time Series

  • 章节

  • 发表日期：

    2025年

查看全文
  • Elena-Simona Apostol,
  • Ciprian-Octavian Truică

摘要

The field of Gravitational Wave research has exploded in recent years. Powerful laser interferometers like Advanced Laser Interferometer Gravitational Wave Observatory (LIGO) and Advanced Virgo are now listening for the universe’s most violent events. While these technological marvels have significantly improved the precision of gravitational wave data collection, the data itself is not perfect. Noise can still creep in, potentially leading to misinterpretations. One problematic type of noise in Gravitational Wave data is the glitch. We define a glitch as a noise event that can either masquerade as a real signal from space or degrade the overall quality of the data. In this chapter, we present our current work for the task of detecting Gravitational Wave glitches using Machine Learning and Deep Learning models. We also establish a clear benchmark to compare their performance, highlighting the models that achieve the best results based on three key metrics: accuracy, precision, and recall.