Application of Machine Learning in the Recognition of Submarine Pipeline Signal

By studying the application of machine learning in the recognition of pipeline signal collected by sub-bottom profle, the authors discussed the feasibility of the automatic processing of submarine pipeline data. According to the testing effect of six kinds of machine learning algorithms in the processing of sub-bottom profle dataset, four of them were elected and applied to the actual processing, with the recognition rate up to 95.3%. It's believed that machine learning has a broad application prospect in the feld of geophysical data interpretation.