Fast Bayesian spectral analysis using Convolutional Neural Networks: Applications over GONG H$\alpha$ solar data
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.