 |
| Titel |
EMV data analysis using Bayesian inference |
| VerfasserIn |
Gregory Koulouras, Ilias Kiourktsidis, Evagelos Tasoulas, Dimitris Agiakatsikas, Kyriakos Kontakos, Constantinos Nomicos |
| Konferenz |
EGU General Assembly 2011
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 13 (2011) |
| Datensatznummer |
250049823
|
|
|
|
|
|
| Zusammenfassung |
| 1 Abstract
Several Data Acquisition Systems are collecting Electromagnetic Variation (EMV) data in
several geographic areas where seismic activities are frequently present. The analysis of the
enormous quantity of data is a challenging task. Here is proposed a data analysis mechanism
that can provide inference about seismic activity. Bayesian network (BN) is a statistical
tool that can learn from previous events. Then by observing current data can give
inference about related events. Given the EMVs in a geographical area, the BN
can be used to compute the probability of the presence of seismic activity in this
area.
Bayesian Networks are robust when there are missing input data. In this case, Bayesian
networks will make the best possible inference by using whatever information is presented.
This is possible because BNs encode the correlation between the input variables. The more
information you supply to them, the more accurate the results will be. This is one of the main
advantages over other data analysis representations like rule bases, decision trees and
artificial neural networks; or over other data analysis techniques like regression, clustering,
classification and density estimation. |
|
|
|
|
|
|