Detecting signals in noisy data via Bayesian approaches, machine learning, and neural networks

The objective of this work is to investigate the presence of clean signals in data with a high level of noise. The research focuses on the application of modern statistical methods, such as the Bayesian approach, combined with contemporary machine learning techniques and deep neural networks capable of detecting nonlinear patterns and improving accuracy in high-complexity scenarios. To achieve this goal, synthetic data containing both clean and distorted signals were used, employing spectral, temporal, and wavelet feature extraction techniques. Additionally, classi cation algorithms were applied to evaluate the e ectiveness of the proposed method. The integration of these approaches enabled the development of an analytical system capable of identifying hidden clean signals in noisy data and testing its e ciency under various conditions.