{"references": ["C. Chiaruttini, V. Roberto and F. Saitta, \"Artificial intelligences in\nseismic signal interpretation\" Geophys. J. Int, 98, pp. 223-232, 1989.", "M. Allameh Zadeh and P. Nassery, \"Application of quadratic neural\nnetworks to seismic signal classification\" Phys. Earth. Planetary Interiors, 113,\npp. 103-110, 1999.", "E. Del Pezzo, A. Esposito, F. Giudicepietro, M. Marinaro, M. Martini\nand S. Scarpetta, \"Discrimination of earthquakes and underwater explosions\nusing neural networks\" Bull. Seism. Soc. Am, 93, pp. 215-223, 2003.", "F. Hlawasch and G. F. Boudreaux-Bartels, \"Linear and quadratic timefrequency\nsignal representations\" IEEE Sig. Proc. Mag, 9, pp. 21-67, 1992.", "A. Papandreou-Suppappola, F. Hlawasch and G. Boudreaux-Bartels,\n\"Quadratic time-frequency representations with scale covariance and\ngeneralized time-shift covariance: a unified framework for the affine,\nhyperbolic and power classes\" Digital signal Processing, 8, pp.3-48, 1998.", "L. Cohen, \"Generalized phase-space distribution functions\" J. Math.\nPhys, 7, pp. 781-786, 1966.", "L. Cohen, \"Time-frequency analysis\" Prentice Hall, 1995.", "O. Rioul and P. Flandrin, \"Time-scale energy distributions : a general\nclass extending wavelet transforms\" IEEE Trans on Signal Processing, 40, pp.\n1746-1757, 1992.", "P. Flandrin, \"Temps-fr\u00e9quence\", Academic Press, 1998.\n[10] F. Hlawasch, A. F. Papandreou-Suppappola and G. Boudreaux-Bartels,\n\"The power classes of quadratic time-frequency representations : a\ngeneralization of the hyperbolic and affine classes\" In 27th Asilomar Conf on\nSignals, Systems and computers, Pacific Grove, CA, 1265-1270, 1993.\n[11] F. Hlawasch, A. F. Papandreou-Suppappola and G. Boudreaux-Bartels,\n\"The hyperbolic class of quadratic time-frequency representations. Part II:\nSubclasses, intersection with affine and power classes, regularity unitarity\"\nIEEE Trans on Signal Processing, 45, pp. 303-315, 1997.\n[12] A. Papandreou-Suppappola, F. Hlawasch and G. Boudreaux-Bartels,\n\"Power class time-frequency representations : interference geometry,\nsmoothing and implementation\" In IEEE Symposium on Time-Frequency and\nTime-Scale Analysis, Paris, pp.193-196, 1996.\n[13] I. Daubechies, \"Ten lectures on wavelets\", SIAM, Philadelphia, Pa,\n1992.\n[14] C. Torrence and G. P. Compo, \"A practical guide to wavelet analysis\"\nBull. Amer. Meteor. Soc, 79, pp. 61-78, 1998.\n[15] M. Benbrahim, K. Benjelloun and A. Ibenbrahim, \"Discrimination des\nsignaux sismiques par r\u00e9seaux de neurones artificiels\" In Proc of 3\u00e8mes\njourn\u00e9es nationales sur les syst\u00e8mes intelligents: th\u00e9orie et applications, Rabat,\nMorocco, pp. 62-66; 2004.\n[16] R. Bellman, \"Adaptive control processes: A guided tour\" Princeton\nUniversity Press, Princeton, 1961.\n[17] W.B. Johnson and J. Lindenstrauss, \"Extensions of Lipshitz mapping\ninto Hilbert space\" In Conference in modern analysis and probability, volume\n26 of Contemporary Mathematics, Amer. Math. Soc, pp. 189-206, 1984.\n[18] I. T. Jolliffe, \"Principal component analysis\", Springer-Verlag, 1986.\n[19] J. E. Jackson, \"A user's guide to principal components\", John Wiley,\nNew York, 1991."]}

The automatic discrimination of seismic signals is an important practical goal for earth-science observatories due to the large amount of information that they receive continuously. An essential discrimination task is to allocate the incoming signal to a group associated with the kind of physical phenomena producing it. In this paper, two classes of seismic signals recorded routinely in geophysical laboratory of the National Center for Scientific and Technical Research in Morocco are considered. They correspond to signals associated to local earthquakes and chemical explosions. The approach adopted for the development of an automatic discrimination system is a modular system composed by three blocs: 1) Representation, 2) Dimensionality reduction and 3) Classification. The originality of our work consists in the use of a new wavelet called "modified Mexican hat wavelet" in the representation stage. For the dimensionality reduction, we propose a new algorithm based on the random projection and the principal component analysis.