In spite of the advantages in handling nonlinear problems, the computation efficiency of KPCA is admittedly slow for large training dataset, which is severely impeding its practical applications. Therefore, this paper attempts to boost the calculation efficiency by introducing a sparse kernel skill, which could greatly streamline the training dataset while effectively preserving its representative information. With this method, the calculation efficiency for seismic denoising is raised nearly by 7 times than the traditional KPCA, while a much higher fitting rate (98.81%) on fluid identification is also achieved as well, even with much fewer training nodes. At present, results with this method is sufficiently rewarding and encouraging enough to motivate further study. Probably, this method would bring profound changes on KPCA's theory and applications in geophysical exploration world.


Article metrics loading...

Loading full text...

Full text loading...

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error