Nonlinear Exploratory Data Analysis Applied to Seismic Signals

This paper compares three unsupervised projection methods: Principal
Component Analysis (PCA), which is linear, Self-Organizing Map (SOM) and
Curvilinear Component Analysis (CCA), which are both nonlinear. Performance
comparison of the three methods is made on a set of seismic data recorded on
Stromboli that includes three classes of signals: explosion-quakes, landslides,
and microtremors. The unsupervised analysis of the signals is able to discover
the nature of the seismic events. Our analysis shows that the SOM algorithm discriminates
better than CCA and PCA on the data under examination.

Demartines, P., Herault, J.: Curvilinear Component Analysis: A Self Organizing Neural
Network for Nonlinear Mapping of Data Sets. IEEE Trans. on Neural Networks, Vol. 8
(1997) 48-154
Herault, J., Guerin-Dugue, A., Villemain, P.: Searching for the Embedded Manifolds in Highdimensional
Data, Problems and Unsolved Questions. Proceedings of ECANN Bruge (2002)
Jollife. I.T.:Principal Component Analysis. Springer Verlag, New York (1986)
Kohonen, T., Hynninen, J., Kangas, J., Laaksonen, J.: SOM_PAK Program Package. Report
A31. Helsinki University, Finland (1996)
Kohonen, T.: Self-Organizing Maps. Series in Information Sciences, Vol. 30. Springer,
Heidelberg. Second ed (1997)
Lee, J.A., Lendasse, A., Donckers, N., Verleysen, M.: A Robust Nonlinear Projection
Method. Proceedings of ESANN’2000 D-Facto pubbl. (2000) 13-20
Makhoul, J.: Linear prediction: a Tutorial Review. IEEE Vol. 63 (1975). 561-580.
Esposito, A. M., Giudicepietro, F., Scarpetta, S., D’Auria, L., Marinaro, M., Martini .M.:
Automatic Discrimination of Landslides Seismic Signals at Stromboli Volcano using Neural
Network (submitted)
Masiello, S., Esposito, A. M., Scarpetta, S., Giudicepietro, F., Esposito, A., Marinaro, M.:
Application of Self Organized Maps and Curvilinear Components Analysis to the
Discrimination of Vesuvius Seismic Signals. To appear in the Proceedings of the Workshop
on Self Organizing Map (WSOM), Paris, 5-8 September (2005)