The Oil Layer Recognition Based on Multi-kernel Function Relevance Vector Machines
In the oil layer recognition, Relevance vector machines (RVM) have a good effect. But the single kernel function RVM has some limitations, a kind of multi-kernel function RVM based on particle swarm optimization (PSO) is proposed, which includes the model parameter estimation, model optimization on...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | EJournal Article |
| Published: |
Institute of Advanced Engineering and Science,
2014-02-01.
|
| Subjects: | |
| Online Access: | Get fulltext |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In the oil layer recognition, Relevance vector machines (RVM) have a good effect. But the single kernel function RVM has some limitations, a kind of multi-kernel function RVM based on particle swarm optimization (PSO) is proposed, which includes the model parameter estimation, model optimization on multi-kernel function RVM, PSO-based training, and recognition. The results of simulation experiment for a typical recognition dataset show that its effect is superior to that of classical RVM and PSO-based single kernel function RVM, and its actual application for oil layer recognition in well logging indicates that the recognition results are completely consistent with the conclusions of oil trial, and it has the high recognition accuracy and the good recognition effect. DOI : http://dx.doi.org/10.11591/telkomnika.v12i2.4167 |
|---|