Dans cette Note, nous présentons une estimation des paramètres du mélange de lois basée sur les distances de Wasserstein et de Cramèr–von Mises–Hellinger. L'approche est illustrée par une simulation dans le cas d'un mélange gaussien unidimensionnel.
In this Note, we present an estimation of the parameters of the probability-mixture based upon the Wasserstein and Cramèr–von Mises–Hellinger distances. This approach is illustrated by a simulation in the case of the univariate Gaussian mixture.
Accepté le :
Publié le :
@article{CRMATH_2004__339_9_653_0, author = {Belili, Nacereddine and Heinich, Henri}, title = {Estimation du m\'elange de probabilit\'es}, journal = {Comptes Rendus. Math\'ematique}, pages = {653--658}, publisher = {Elsevier}, volume = {339}, number = {9}, year = {2004}, doi = {10.1016/j.crma.2004.09.019}, language = {fr}, url = {http://archive.numdam.org/articles/10.1016/j.crma.2004.09.019/} }
TY - JOUR AU - Belili, Nacereddine AU - Heinich, Henri TI - Estimation du mélange de probabilités JO - Comptes Rendus. Mathématique PY - 2004 SP - 653 EP - 658 VL - 339 IS - 9 PB - Elsevier UR - http://archive.numdam.org/articles/10.1016/j.crma.2004.09.019/ DO - 10.1016/j.crma.2004.09.019 LA - fr ID - CRMATH_2004__339_9_653_0 ER -
%0 Journal Article %A Belili, Nacereddine %A Heinich, Henri %T Estimation du mélange de probabilités %J Comptes Rendus. Mathématique %D 2004 %P 653-658 %V 339 %N 9 %I Elsevier %U http://archive.numdam.org/articles/10.1016/j.crma.2004.09.019/ %R 10.1016/j.crma.2004.09.019 %G fr %F CRMATH_2004__339_9_653_0
Belili, Nacereddine; Heinich, Henri. Estimation du mélange de probabilités. Comptes Rendus. Mathématique, Tome 339 (2004) no. 9, pp. 653-658. doi : 10.1016/j.crma.2004.09.019. http://archive.numdam.org/articles/10.1016/j.crma.2004.09.019/
[1] Application of EM-type algorithms to spatial data, Commun. Statist. Theory Methods, Volume 26 (1997) no. 3, pp. 669-683
[2] Statistical analysis of mixtures and the empirical probability measure, Acta Appl. Math., Volume 50 (1998) no. 3, pp. 253-340
[3] Transport problem and derivation, Appl. Math., Volume 26 (1999) no. 3, pp. 299-314
[4] Estimations basées sur la fonctionnelle de Kantorovich et la distance de Lévy, C. R. Acad. Sci. Paris, Ser. I., Volume 328 (1999), pp. 423-426
[5] Some asymptotic theory for boostrap, Ann. Statist., Volume 9 (1981), pp. 1196-1217
[6] Reconnaissance de mélanges de densités par un algorithme d'apprentissage probabiliste, Data Anal. Inform, Volume 3 (1983), pp. 359-374
[7] G. Celeux, S. Chrétien, A. Mkhadri, A component-wise EM algorithm for mixtures, Research Report INRIA, 3746, 1999
[8] Mass transportation problems in probability theory, Math. Scientist, Volume 21 (1996), pp. 34-72
[9] Convergence of a stochastic approximation version of the EM algorithm, Ann. Statist., Volume 27 (1999) no. 1, pp. 94-128
[10] Asymptotic properties of a stochastic EM algorithm for estimating mixing proportions, Commun. Statist. Stochastic Models, Volume 9 (1993) no. 4, pp. 599-613
[11] Algorithmes Stochastiques, Math. Appl., vol. 23, SMAI, Springer, 1996
[12] Computing the observed information in the hidden Markov model using the EM algorithm, Statist. Probab. Lett., Volume 32 (1997) no. 1, pp. 107-114
[13] The EM Algorithm and Extensions, Wiley Ser. Probab. Statist., Wiley, New York, 1997
[14] Mass Transportation Problems, Springer, New York, 1998
Cité par Sources :