Tecniche Soft-computing per il calcolo della costante dielettrica delle ceneri vulcaniche
Conference Paper
Publication Date:
2005
Short description:
Tecniche Soft-computing per il calcolo della costante dielettrica delle ceneri vulcaniche / Barrile, Vincenzo; M., Cacciola. - (2005). (Intervento presentato al convegno Convegno Nazionale SIFET - 2005 tenutosi a Mondello (PA) nel 29 giugno - 1 luglio 2005).
abstract:
The resolution of many problems inherent the treatment and the analysis of acquired data with most different techniques and
methods of "survey" mainly require determination of new values of the signal. This phase, known and subsequently denoted with the
terminology of "prediction of the signal", can follows the separation in relief data of a correlated part called "signal" from a
incorrelated part known as "noise". Already in [ Barrile, Cacciola, 2004] we theorically treated about am inferential model which is
flexible, robustness with improved performances as regards the classical stochastic systems. In this work we test its value by the
electromagnetic characterization of the volcanic ashes, which is very important as concerning the aerial safety, as for the
predictibility of possible volcanic eruptions. Comparing the Round Mean Square Error (RMSE) and the RMSE Percent (RMSEP)
values between the predicted and the empirical values, we will show the goodness of a heuristic system of filtering and, above all, of
prediction, by comparison of different "intelligent" structures, which are reciprocally interchangeable according to the application.
This application study is involved inner a more complex research project: the aim is monitoring and controlling ground aeroportual
traffic by integrated techniques using GPS and GIS. At the same time, this application is involved in another research project which
aim is to identify, characterize and control volcano eruptions, mainly for civil purposes.
methods of "survey" mainly require determination of new values of the signal. This phase, known and subsequently denoted with the
terminology of "prediction of the signal", can follows the separation in relief data of a correlated part called "signal" from a
incorrelated part known as "noise". Already in [ Barrile, Cacciola, 2004] we theorically treated about am inferential model which is
flexible, robustness with improved performances as regards the classical stochastic systems. In this work we test its value by the
electromagnetic characterization of the volcanic ashes, which is very important as concerning the aerial safety, as for the
predictibility of possible volcanic eruptions. Comparing the Round Mean Square Error (RMSE) and the RMSE Percent (RMSEP)
values between the predicted and the empirical values, we will show the goodness of a heuristic system of filtering and, above all, of
prediction, by comparison of different "intelligent" structures, which are reciprocally interchangeable according to the application.
This application study is involved inner a more complex research project: the aim is monitoring and controlling ground aeroportual
traffic by integrated techniques using GPS and GIS. At the same time, this application is involved in another research project which
aim is to identify, characterize and control volcano eruptions, mainly for civil purposes.
Iris type:
4.1 Contributo in Atti di convegno
Keywords:
Ceneri vulcaniche, costante dielettrica, caratteristiche elettromagnetiche, predizione, sistemi esperti
List of contributors:
Barrile, Vincenzo; M., Cacciola
Book title:
SIFET