标题： 使用卷积神经网络的快速贝叶斯光谱分析：GONG H$α$太阳数据的应用 显示英文标题

标题： Fast Bayesian spectral analysis using Convolutional Neural Networks: Applications over GONG H$α$ solar data

摘要： 背景。 太阳丝状体振荡已被观测多年，但望远镜能力的最新进展现在使得这些周期性运动的每日监测成为可能，为丝状体结构提供了宝贵的见解。 对整个太阳周期的丝状体振荡进行系统研究可以揭示日珥的演化。 直到现在，只有手动技术被用来分析这些振荡。 目的。 这项工作旨在证明概念，目的是展示卷积神经网络（CNN）的有效性。 这些网络通过应用功率谱分析到GONG望远镜网络的H$\alpha$数据来自动检测丝状体振荡。 方法。 所提出的技术研究了H$\alpha$数据立方体中每个像素的周期波动。 使用Lomb Scargle周期图，我们计算了数据集的功率谱密度（PSD）。 背景噪声很好地拟合为红噪声和白噪声的组合。 利用贝叶斯统计和马尔可夫链-蒙特卡洛（MCMC）算法，我们可以拟合光谱并确定给定百分比的置信阈值以搜索真实的振荡。 我们构建了两个CNN模型，以获得与MCMC方法相同的结果。 结果。 我们将CNN模型应用于文献中报道的一些观测结果，证明了其在检测与经典方法相同事件方面的可靠性。 研究了一天中之前未报告的事件，以检查模型在非受控数据集中的能力，在该数据集中我们可以与之前的报告进行核对。 结论。 CNN被证明是研究太阳丝状体振荡的一个有用的工具，采用光谱技术。 计算时间显著减少，同时得到的结果与经典方法足够相似。 这是朝着自动检测丝状体振荡迈出的重要一步。 显示英文摘要

摘要： Context. Solar filament oscillations have been observed for many years, but recent advances in telescope capabilities now enable daily monitoring of these periodic motions, offering valuable insights into the structure of filaments. A systematic study of filament oscillations over the solar cycle can shed light on the evolution of the prominences. Until now, only manual techniques have been used to analyze these oscillations. Aims. This work serves as a proof of concept, aiming to demonstrate the effectiveness of Convolutional Neural Networks (CNNs). These networks automatically detect filament oscillations by applying power-spectrum analysis to H$\alpha$ data from the GONG telescope network. Methods. The proposed technique studies periodic fluctuations of every pixel of the H$\alpha$ data cubes. Using the Lomb Scargle periodogram, we computed the power spectral density (PSD) of the dataset. The background noise is well fitted to a combination of red and white noise. Using Bayesian statistics and Markov chains -- Monte Carlo (MCMC) algorithms we can fit the spectra and determine the confidence threshold of a given percentage to search for real oscillations. We built two CNN models to obtain the same results as the MCMC approach. Results. We applied the CNN models to some observations reported in the literature, proving its reliability in detecting the same events as the classical methods. A day with events not previously reported has been studied to check for the model capabilities outside of a controlled dataset where we can check with previous reports. Conclusions. CNNs proved to be a useful tool to study solar filament oscillations, using spectral techniques. Computing times have been reduced significantly while getting results similar enough to the classical methods. This is a relevant step towards the automatic detection of filament oscillations.