Pattern Recognition for Flank Eruption Forecasting: An Application at Mount Etna Volcano Sicily, ItalyReportar como inadecuado




Pattern Recognition for Flank Eruption Forecasting: An Application at Mount Etna Volcano Sicily, Italy - Descarga este documento en PDF. Documentación en PDF para descargar gratis. Disponible también para leer online.

Avolcano can be defined as a complex system, not least for the hidden cluesrelated to its internal nature. Innovative models grounded in the ArtificialSciences, have been proposed for a novel pattern recognition analysis at Mt.Etna volcano. The reference monitoring dataset dealt with real data of 28parameters collected between January 2001 and April 2005, during which thevolcano underwent the July-August 2001, October 2002-January 2003 and September2004-April 2005 flank eruptions. There were 301 eruptive days out of an overallnumber of 1581 investigated days. The analysis involved successive steps.First, the TWIST algorithm was used to select the most predictive attributesassociated with the flank eruption target. During his work, the algorithm TWISTselected 11 characteristics of the input vector: among them SO2 andCO2 emissions, and also many other attributes whose linearcorrelation with the target was very low. A 5 × 2 Cross Validation protocolestimated the sensitivity and specificity of pattern recognition algorithms.Finally, different classification algorithms have been compared to understandif this pattern recognition task may have suitable results and which algorithmperforms best. Best results higher than 97% accuracy have been obtained afterperforming advanced Artificial Neural Networks, with a sensitivity andspecificity estimates over 97% and 98%, respectively. The present analysishighlights that a suitable monitoring dataset inferred hidden information aboutvolcanic phenomena, whose highly non-linear processes are enhanced.

KEYWORDS

Mt. Etna Volcano, Flank Eruption Forecasting, Neural Networks, Pattern Recognition, Monitoring Data

Cite this paper

Brancato, A. , Buscema, P. , Massini, G. and Gresta, S. 2016 Pattern Recognition for Flank Eruption Forecasting: An Application at Mount Etna Volcano Sicily, Italy. Open Journal of Geology, 6, 583-597. doi: 10.4236-ojg.2016.67046.





Autor: A. Brancato1*, P. M. Buscema2,3, G. Massini2, S. Gresta4

Fuente: http://www.scirp.org/



DESCARGAR PDF




Documentos relacionados