|
Earth-prints >
Editorial Initiatives >
eJournals >
Annals of Geophysics >
Please use this identifier to cite or link to this item:
http://hdl.handle.net/2122/1874
|
| Title: | Seismic signals detection and classification using artiricial neural networks |
| Authors: | Romeo, G.* |
| Keywords: | seismology
detection
neural network
auto-associative neural network
classification |
| Issue Date: | Jun-1994 |
| Series/Report no.: | 37/3 |
| Abstract: | Pattern recognition belongs to a class of Problems which are easily solved by humans, but difficult for computers. It is sometimes difficult to formalize a problem which a human operator can casily understand by using examples. Neural networks are useful in solving this kind of problem. A neural network may, under certain conditions, simulate a well trained human operator in recognizing different types of earthquakes or in detecting the presence of a seismic event. It is then shown how a fully connected multi layer perceptron may perform a recognition task. It is shown how a self training auto associative neural network may detect an earthquake occurrence analysing the change in signal characteristics. |
| URI: | http://hdl.handle.net/2122/1874 |
| Appears in Collections: | Annals of Geophysics
05.01.99. General or miscellaneous
|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.
|
Del.icio.us
Citeulike
Connotea
Facebook
Stumble it!
reddit
