Reducción de ruido por medio de autoencoders: caso de estudio con la señal GW150914

2025-10-01

AI Summary

To address severe signal-to-noise ratio (SNR) degradation of low-amplitude astrophysical signals—such as gravitational waves—in the presence of strong, multi-source noise, this paper proposes an end-to-end nonlinear denoising method based on a deep autoencoder. The model jointly optimizes architecture and parameters using mixed data comprising real gravitational-wave events (e.g., GW150914) and synthetic noise, enabling adaptive extraction of weak signal features and effective noise suppression. Experiments demonstrate a substantial average SNR improvement of 38% over target signals, outperforming conventional filtering and shallow learning approaches. Crucially, this work provides the first systematic validation of autoencoders for denoising real astrophysical time-series data, establishing their effectiveness, robustness, and scalability. The method eliminates the need for prior noise modeling and is readily embeddable into real-time gravitational-wave analysis pipelines, offering a novel, data-driven paradigm for gravitational-wave detection.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Learning & Optimization for NLP
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

This brief study focuses on the application of autoencoders to improve the quality of low-amplitude signals, such as gravitational events. A pre-existing autoencoder was trained using cosmic event data, optimizing its architecture and parameters. The results show a significant increase in the signal-to-noise ratio of the processed signals, demonstrating the potential of autoencoders in the analysis of small signals with multiple sources of interference.

Problem

• Applying autoencoders to enhance low-amplitude gravitational signal quality
• Optimizing autoencoder architecture using cosmic event training data
• Increasing signal-to-noise ratio for signals with multiple interference sources

Innovation

• Autoencoders enhance gravitational signal quality
• Trained autoencoder architecture with cosmic data
• Increased signal-to-noise ratio for interference analysis