Multi-Layer Perceptions in Filtering Geophysical Signals

Multi-layer neural networks are used to model the influence of noise sources on some type of geophysicalsignals. The role of neural networks is to identify non-linear time invariant NARMAX models which are used to filter corrupted data using a model-based approach, avoiding the drawbacks of the classical digital filtering techniques. In this paper the goodness of the proposed approach is shown by considering a case study concerning with the filtering of tilt signals recorded in volcanic active areas.