Reconnaissance supervisée et non supervisée de lois à partir d'échantillons finis

Dans cet article, nous abordons le probleme de la reconnaissance de lois de probabilite a partir d'echantillons variant de 100 a 10 000 ou plus. Le contexte applicatif porte sur la modelisation de canaux radio-mobile en situation de visibilite ou de non-visibilite directe entre emetteur et recepteur. Ce probleme est crucial pour ameliorer les communications numeriques. Dans la communaute des transmissions numeriques, il est courant d'utiliser la distance de Kolmogorov-Smirnov. Plus rarement, une methode a noyau est consideree avant le test comparatif. Nous proposons d'utiliser les criteres d'information (IC), d'une part pour approcher les lois de probabilite par un histogramme, et d'autre part pour selectionner le meilleur modele de loi. Nous etudions les cas supervise et non supervise et comparons les methodes dans ces situations realistes. Les resultats montrent l'interet d'utiliser les methodes exploitant les IC.

[1]  Olivier Alata,et al.  Reconnaissance de lois via une estimation de densité par histogramme pour la modélisation de canaux de transmission , 2009 .

[2]  Guilhem Coq.  Utilisation d'approches probabilistes basées sur les critères entropiques pour la recherche d'information sur supports multimédia , 2008 .

[3]  Zhengqing Yun,et al.  Propagation prediction models for wireless communication systems , 2002 .

[4]  H. Hashemi,et al.  Statistical modeling and simulation of the RMS delay spread of indoor radio propagation channels , 1994 .

[5]  R. Shibata.  Selection of the order of an autoregressive model by Akaike's information criterion , 1976 .

[6]  Christian Olivier,et al.  Law recognition via histogram-based estimation , 2009, 2009 IEEE International Conference on Acoustics, Speech and Signal Processing.

[7]  Jorma Rissanen,et al.  Density estimation by stochastic complexity , 1992, IEEE Trans. Inf. Theory.

[8]  Clifford M. Hurvich,et al.  Regression and time series model selection in small samples , 1989 .

[9]  Marc Hallin,et al.  Order Selection, Stochastic Complexity and Kullback-Leibler Information , 1996 .

[10]  Christian Olivier,et al.  Information criteria based edge detection , 1998, 9th European Signal Processing Conference (EUSIPCO 1998).

[11]  Piet M. T. Broersen,et al.  Finite sample criteria for autoregressive order selection , 2000, IEEE Trans. Signal Process..

[12]  Zdravko I. Botev,et al.  A Novel Nonparametric Density Estimator , 2006 .

[13]  M. Salazar-Palma,et al.  A survey of various propagation models for mobile communication , 2003 .

[14]  G. Schwarz.  Estimating the Dimension of a Model , 1978 .

[15]  B. G. Quinn,et al.  The determination of the order of an autoregression , 1979 .

[16]  Fulvio Babich,et al.  Statistical analysis and characterization of the indoor propagation channel , 2000, IEEE Trans. Commun..

[17]  Peter Grünwald,et al.  A tutorial introduction to the minimum description length principle , 2004, ArXiv.

[18]  Olivier Alata,et al.  Information criteria and arithmetic codings : An illustration on raw images , 2007, 2007 15th European Signal Processing Conference.

[19]  Deniz Erdogmus,et al.  Entropy minimization for supervised digital communications channel equalization , 2002, IEEE Trans. Signal Process..

[20]  J. Rissanen,et al.  Modeling By Shortest Data Description* , 1978, Autom..

[21]  H. Akaike.  A new look at the statistical model identification , 1974 .

[22]  Bouchra Senadji,et al.  A statistical model for the simulation of time-varying multipath mobile radio propagation channel , 1994, Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing.

[23]  M. Basseville.  Distance measures for signal processing and pattern recognition , 1989 .

[24]  Abbas Jamalipour,et al.  Wireless communications , 2005, GLOBECOM '05. IEEE Global Telecommunications Conference, 2005..

[25]  Jorma Rissanen,et al.  Generalized Kraft Inequality and Arithmetic Coding , 1976, IBM J. Res. Dev..

[26]  Aawatif Hayar,et al.  A performance study of Kullback-Leibler distance-based spectrum sensing algorithm , 2011, 2011 3rd International Congress on Ultra Modern Telecommunications and Control Systems and Workshops (ICUMT).

[27]  R. J. Davies,et al.  Propagation considerations for the design of an indoor broad-band communications system at EHF , 1998 .