Identification and classification of multiple reflections with self-organizing maps1

Artificial neural networks can be used effectively to identify and classify multiple events in a seismic data set. We use a specialized neural network, a self-organizing map, that tries to establish rules for the characterisation of the physical problem. Selected seismic data attributes are used as input patterns, such that the self-organizing map arranges the data in a manner that forms clusters in an abstract space. We show with synthetic and real data how the artificial neural network can to identify and classify primaries, multiples, and how it can classify the various types of multiples corresponding to a certain generating mechanism in the subsurface.