Home | Current | Past volumes | About | Login | Notify | Contact | Search
 Statistics Surveys > Vol. 3 (2009) open journal systems 


Distributional properties of means of random probability measures

Antonio Lijoi, University of Pavia
Igor Prünster, University of Turin


Abstract
The present paper provides a review of the results concerning distributional properties of means of random probability measures. Our interest in this topic has originated from inferential problems in Bayesian Nonparametrics. Nonetheless, it is worth noting that these random quantities play an important role in seemingly unrelated areas of research. In fact, there is a wealth of contributions both in the statistics and in the probability literature that we try to summarize in a unified framework. Particular attention is devoted to means of the Dirichlet process given the relevance of the Dirichlet process in Bayesian Nonparametrics. We then present a number of recent contributions concerning means of more general random probability measures and highlight connections with the moment problem, combinatorics, special functions, excursions of stochastic processes and statistical physics.

AMS 2000 subject classifications: Primary 62F15, 62E15; secondary 60G57.

Keywords: Bayesian Nonparametrics, completely random measures, Cifarelli–Regazzini identity, Dirichlet process, functionals of random probability measures, generalized Stieltjes transform, neutral to the right processes, normalized random measures, posterior distribution, random means, randomprobabilitymeasure, two–parameter Poisson–Dirichlet process.

Creative Common LOGO

Full Text: PDF


Lijoi, Antonio, Prünster, Igor, Distributional properties of means of random probability measures, Statistics Surveys, 3, (2009), 47-95 (electronic). DOI: 10.1214/09-SS041.

References

[1]     Barlow, M., Pitman, J. and Yor, M. (1989). Une extensionBarlow multidimensionnelle de la loi de l’arc sinus. In Séminaire de Probabilités XXIII (Azema, J., Meyer, P.-A. and Yor, M., Eds.), 294–314, Lecture Notes in Mathematics 1372. Springer, Berlin. MR1022893

[2]     Bertoin, J. and Yor, M. (2005). Exponential functionals of LévyBertoinYor processes. Prob. Surveys 2, 191–212. MR2178044

[3]     Blackwell, D. (1973). Discreteness of Ferguson selections. Ann.black Statist. 1, 356-358. MR0348905

[4]     Brix, A. (1999). Generalized gamma measures and shot-noise Coxbrix processes. Adv. Appl. Probab. 31, 929–953. MR1747450

[5]     Carlton, M.A. (2002). A family of densities derived fromCarlton the three-parameter Dirichlet process. J. Appl. Probab. 39, 764–774. MR1938169

[6]     Carmona, P., Petit, F. and Yor, M. (1997). On the distributioncarmpetityor and asymptotic results for exponential functionals of Lévy processes. In Exponential functionals and principal values related to Brownian motion (M. Yor, ed.), 73–126. Biblioteca de la Revista Matematica Ibero–Americana, Madrid. MR1648657

[7]     Cifarelli, D.M. and Melilli, E. (2000). Some new results forpieveloce Dirichlet priors. Ann. Statist. 28, 1390–1413. MR1805789

[8]     Cifarelli, D.M. and Regazzini, E. (1979a). A general approach tocifreg79a Bayesian analysis of nonparametric problems. The associative mean values within the framework of the Dirichlet process. I. (Italian) Riv. Mat. Sci. Econom. Social. 2, 39–52. MR0573686

[9]     Cifarelli, D.M. and Regazzini, E. (1979b). A general approach tocifreg79b Bayesian analysis of nonparametric problems. The associative mean values within the framework of the Dirichlet process. II. (Italian) Riv. Mat. Sci. Econom. Social. 2, 95–111. MR0574040

[10]     Cifarelli, D.M. and Regazzini, E. (1990). Distribution functionscifreg90 of means of a Dirichlet process. Ann. Statist., 18, 429–442 (Correction in Ann. Statist. (1994) 22, 1633–1634). MR1041402

[11]     Cifarelli, D.M. and Regazzini, E. (1993). Some remarks on thecifreg93 distribution functions of means of a Dirichlet process. Technical Report 93.4, IMATI–CNR, Milano.

[12]     Cifarelli, D.M. and Regazzini, E. (1996). Tail-behaviour and finitenesscifreg96 of means of distributions chosen from a Dirichlet process. Technical Report 96.19, IMATI–CNR, Milano. Available at: http://www.mi.imati.cnr.it/ iami/papers/96-19.ps

[13]     Constantine, G.M. and Savits, T.H. (1996). A multivariateconst.savits Faà di Bruno formula with applications. Trans. Amer. Math. Soc. 348, 503–520. MR1325915

[14]     Daley, D.J. and Vere–Jones, D. (1988). An introduction to thedaley theory of point processes. Springer, New York. MR0950166

[15]     Diaconis, P. and Kemperman, J. (1996). Some new tools ondiackemp Dirichlet priors. In Bayesian statistics 5 (J.M. Bernardo, J.O. Berger, A.P. Dawid and A.F.M. Smith, eds.) 97–106. Oxford Univ. Press. MR1425401

[16]     Dickey, J.M. and Jiang, T.J. (1998). Filtered-variate priordickey distributions for histogram smoothing. J. Amer. Statist. Assoc. 93, 651–662. MR1631349

[17]     Doksum, K. (1974). Tailfree and neutral random probabilities anddok their posterior distributions. Ann. Probab. 2, 183–201. MR0373081

[18]     Dykstra, R.L. and Laud, P. (1981). A Bayesian nonparametricdykstra approach to reliability. Ann. Statist. 9, 356–367. MR0606619

[19]     Epifani, I., Guglielmi, A. and Melilli, E. (2006). A stochasticepifani equation for the law of the random Dirichlet variance. Statist. Probab. Lett. 76, 495–502. MR2266603

[20]     Epifani, I., Lijoi, A. and Prünster, I. (2003). Exponentialelp functionals and means of neutral–to–the–right priors. Biometrika 90, 791–808. MR2024758

[21]     Erhardsson, T. (2008). Nonparametric Bayesian inference forErh(08) integrals with respect to an unknown finite Mmeasure. Scand. J. Statist. 35, 354–368. MR2418747

[22]     Evans, M.R. and Hanney, T. (2005). Nonequilibrium statisticalevans mechanics of the zero-range process and related models. J. Phys. A 38, R195-R240. MR2145800

[23]     Exton, H. (1976). Multiple hypergeometric functions and applications.exton Ellis Horwood, Chichester. MR0422713

[24]     Favaro, S., Guglielmi, A. and Walker, S.G. (2009). A classfavaro of measure-valued Markov Chains and Bayesian Nonparametrics. Tech. Report.

[25]     Feigin, P.D. and Tweedie, R.L. (1989). Linear functionals andFeigTweed Markov chains associated with Dirichlet processes. Math. Proc. Camb. Phil. Soc. 105, 579–585. MR0985694

[26]     Ferguson, T.S. (1973). A Bayesian analysis of some nonparametricferg2 problems. Ann. Statist. 1, 209–230. MR0350949

[27]     Ferguson, T.S. (1974). Prior distributions on spaces of probabilityferg3 measures. Ann. Statist. 2, 615–629. MR0438568

[28]     Guglielmi, A. (1998). A simple procedure calculating the generalizedGug(98) Stieltjes transform of the mean of a Dirichlet process. Statist. Probab. Lett. 38, 299–303. MR1631205

[29]     Guglielmi, A. and Tweedie, R.L. (2001). Markov chain MonteGuglTweed Carlo estimation of the law of the mean of a Dirichlet process. Bernoulli 7, 573–592. MR1849368

[30]     Guglielmi, A., Holmes, C.C. and Walker, S.G. (2002). PerfectGuglHolmesWalker simulation involving functionals of a Dirichlet process. J. Comput. Graph. Statist. 11, 306–310. MR1938137

[31]     Gurland, J. (1948). Inversion formulae for the distributions of ratios.gurl Ann. Math. Statist., 19, 228–237. MR0025005

[32]     Haas, B., Miermont, G., Pitman, J. and Winkel, M. (2008).Haas Continuum tree asymptotics of discrete fragmentations and applications to phylogenetic models. Ann. Probab. 36, 1790–1837. MR2440924

[33]     Hannum, R.C., Hollander, M. and Langberg, N.A. (1981).hhl Distributional results for random functionals of a Dirichlet process. Ann. Probab. 9, 665–670. MR0630318

[34]     Hill, T. and Monticino, M. (1998). Constructions of randomMonticino distributions via sequential barycenters. Ann. Statist. 26, 1242–1253. MR1647649

[35]     Hjort, N.L. (2003). Topics in non-parametric Bayesian statistics.cervohsss In: Highly structured stochastic systems (P. Green, N.L. Hjort and S. Richardson, Eds.), 455–487. Oxford Univ. Press, Oxford. MR2082419

[36]     Hjort, N.L. and Ongaro, A. (2005). Exact inference for randomcervongaro Dirichlet means. Stat. Inference Stoch. Process. 8, 227–254. MR2177313

[37]     Hjort, N.L. and Petrone, S. (2007). Nonparametric quantilecervosonia inference using Dirichlet processes. In Advances in statistical modeling and inference, 463–492, World Sci. Publ., Hackensack, NJ. MR2416129

[38]     Ishwaran, H. and James, L.F. (2003). Generalized weighted Chineseishwjames2 restaurant processes for species sampling mixture models. Statist. Sinica 13, 1211–1235. MR2026070

[39]     Ismail, M.E.H. and Kelker, D.H. (1979). Special functions, Stieltjesismailkelker transforms and infinite divisibility. SIAM J. Math. Anal. 10, 884–901. MR0541088

[40]     James, L.F. (2002). Poisson process partition calculus withpaperone applications to exchangeable models and Bayesian nonparametrics. Mathematics ArXiv, math.PR/0205093.

[41]     James, L.F. (2006). Functionals of Dirichlet processes, thelancill2 Cifarelli-Regazzini identity and beta-gamma processes. Ann. Statist. 33, 647–660. MR2163155

[42]     James, L.F. (2007). New Dirichlet mean identities. Mathematicslancill3 Arxiv, math.PR/0708.0614.

[43]     James, L.F., Lijoi, A. and Prünster, I. (2006). Conjugacy as ajlp1 distinctive feature of the Dirichlet process. Scand. J. Statist. 33, 105–120. MR2255112

[44]     James, L.F., Lijoi, A. and Prünster, I. (2008). Distributions ofjlp2 linear functionals of two parameter Poisson–Dirichlet random measures. Ann. Appl. Probab. 18, 521–551. MR2398765

[45]     James, L.F., Lijoi, A. and Prünster, I. (2009). Posterior analysisjlp0 for normalized random measures with independent increments. Scand. J. Statist. 36, 76–97. MR2508332

[46]     James, L.F., Lijoi, A. and Prünster, I. (2009). On thejlp3 posterior distribution of classes of random means. Bernoulli, DOI: 10.3150/09-BEJ200.

[47]     James, L.F., Roynette, B. and Yor, M. (2008). Generalizedjry gamma convolutions, Dirichlet means, Thorin measures, with explicit examples. Probab. Surv. 5, 346–415. MR2476736

[48]     Jarner, S.F., Tweedie, R.L. (2002). Convergence rates andJar(02) moments of Markov chains associated with the mean of Dirichlet processes.Stochastic Process. Appl. 101, 257–271. MR1931269

[49]     Kerov, S.V. (1998). Interlacing measures. In Kirillov’s seminar onKerov representation theory. Amer. Math. Soc. Transl. Ser. 2, Vol. 181. Amer. Math. Soc., Providence, 35–83. MR1618739

[50]     Kerov, S.V. and Tsilevich, N.V. (2001). The Markov-KreinKT correspondence in several dimensions. Zap. Nauchn. Sem. S.-Peterburg. Otdel. Mat. Inst. Steklov. (POMI) 283, 98–122. (English translation in J. Math. Sci. 121, 2345–2359). MR1879065

[51]     Kingman, J.F.C. (1967). Completely random measures. Pacific J.jfc1 Math. 21, 59–78. MR0210185

[52]     Kingman, J.F.C. (1975). Random discrete distributions (withjfc2 discussion). J. Roy. Statist. Soc. Ser. B 37, 1–22. MR0368264

[53]     Kingman, J.F.C. (1993). Poisson processes. Oxford University Press,jfc3 Oxford. MR1207584

[54]     Lamperti, J. (1958). An occupation time theorem for a class ofLamperti stochastic processes. Trans. Amer. Math. Soc. 88, 380–387. MR0094863

[55]     Lévy, P. (1939). Sur certains processus stochastiques homogènes.Levy Compositio Math. 7, 283–339. MR0000919

[56]     Lijoi, A., Mena, R.H. and Prünster, I. (2005). Hierarchicallmp1 mixture modelling with normalized inverse Gaussian priors. J. Amer. Statist. Assoc. 100, 1278–1291. MR2236441

[57]     Lijoi, A., Mena, R.H. and Prünster, I. (2005). Bayesianlmp3 nonparametric analysis for a generalized Dirichlet process prior. Statist. Inference Stoch. Process., 8, 283–309. MR2177315

[58]     Lijoi, A., Mena, R.H. and Prünster, I. (2007). Controlling thelmp2 reinforcement in Bayesian nonparametric mixture models. J. Roy. Statist. Soc. Ser. B 69, 715–740. MR2370077

[59]     Lijoi, A. and Regazzini, E. (2004). Means of a Dirichlet processlr and multiple hypergeometric functions. Ann. Probab. 32, 1469–1495. MR2060305

[60]     Lo, A.Y. (1984). On a class of Bayesian nonparametric estimates. I.lo Density estimates. Ann. Statist. 12, 351–357. MR0733519

[61]     Lo, A.Y. and Weng, C.-S. (1989). On a class of Bayesianloweng nonparametric estimates. II. Hazard rate estimates. Ann. Inst. Statist. Math. 41, 227–245. MR1006487

[62]     Muliere, P. and Tardella, L. (1998). Approximating distributionsmt of random functionals of Ferguson-Dirichlet priors. Canad. J. Statist. 26 283–297. MR1648431

[63]     Nieto-Barajas, L.E., Prünster, I. and Walker, S.G. (2004).npw Normalized random measures driven by increasing additive processes. Ann. Statist. 32, 2343–2360. MR2153987

[64]     Peccati, G. (2008). Multiple integral representation for functionalspeccati of Dirichlet processes. Bernoulli 14, 91–124. MR2401655

[65]     Pitman, J. (1995). Exchangeable and partially exchangeable randomPitman1995 partitions. Probab. Theory Related Fields. 102, 145–158. MR1337249

[66]     Pitman, J. (1996). Some developments of the Blackwell-MacQueenPitman1996 urn scheme. In Statistics, Probability and Game Theory (T.S. Ferguson, L.S. Shapley and J.B. MacQueen, Eds.). IMS Lecture Notes Monogr. Ser., Vol. 30. Inst. Math. Statist., Hayward, 245–267. MR1481784

[67]     Pitman, J. (2003). Poisson-Kingman partitions. In Statistics andpk science: a Festschrift for Terry Speed (D.R Goldstein, Ed.). IMS Lecture Notes Monogr. Ser., Vol. 40. Inst. Math. Statist., Beachwood, 1–34. MR2004330

[68]     Pitman, J. and Yor, M. (1992). Arcsine laws and interval partitionsPitmanYor1992 derived from a stable subordinator. Proc. London Math. Soc. 65 326–356. MR1168191

[69]     Pitman, J., and Yor, M. (1997). On the relative lengths of excursionsPitmanYor1997 derived from a stable subordinator. In: Séminaire de Probabilités XXXI (Azema, J., Emery, M. and Yor, M., Eds.), 287–305, Lecture Notes in Mathematics 1655. Springer, Berlin. MR1478738

[70]     Pulkkinen, O. (2007). Boundary driven zero-range processes inpulkk random media. J. Stat. Phys. 128, 1289–1305. MR2352010

[71]     Regazzini, E. (1998). An example of the interplay between statisticsreglincei and special functions. In Atti dei convegni Lincei. Tricomi’s ideas and contemporary applied mathematics 147, 303–320. MR1737502

[72]     Regazzini, E., Guglielmi, A. and Di Nunno, G. (2002). Theoryrgd and numerical analysis for exact distribution of functionals of a Dirichlet process. Ann. Statist. 30, 1376–1411. MR1936323

[73]     Regazzini, E., Lijoi, A. and Prünster, I. (2003). Distributionalrlp results for means of random measures with independent increments. Ann. Statist. 31, 560–585. MR1983542

[74]     Regazzini, E. and Sazonov, V.V. (2001). Approximation ofregsaz laws of random probabilities by mixtures of Dirichlet distributions with applications to nonparametric Bayesian inference. Th. Prob. Appl., 45, 93–110. MR1810976

[75]     Romik, D. (2004). Explicit formulas for hook walks on continual Youngromik04 diagrams. Adv. in Appl. Math. 32, 625–654. MR2053837

[76]     Romik, D. (2005). Roots of the derivative of a polynomial. Amer.romik05 Math. Monthly 112, 66–69. MR2110114

[77]     Sato, K. (1999). Lévy processes and infinitely divisible distributions.sato Cambridge University Press, Cambridge. MR1739520

[78]     Sethuraman, J. (1994). A constructive definition of the Dirichletseth process prior. Statist. Sinica 2, 639–650. MR1309433

[79]     Schwarz, J.H. (2005) The generalized Stieltjes transform and itsSchwarz inverse. J. Math. Phys. 46 013501-8. MR2113770

[80]     Spitzer, F. (1970). Interaction of Markov processes. Advances inspitzer Math. 5, 246-290. MR0268959

[81]     Stanley, R.P. (1999). Enumerative combinatorics. Volume 2.stanley Cambridge University Press. MR1676282

[82]     Sumner, D.B. (1949). An inversion formula for the generalizedsumner Stieltjes transform. Bull. Amer. Math. Soc. 55, 174–183. MR0027888

[83]     Tamura, H. (1988). Estimation of rare errors using expert judgement.tamura Biometrika 75, 1–9. MR0932811

[84]     Tsilevich, N.V. (1997). Distribution of mean values for sometsil random measures. Zap. Nauchn. Sem. S.-Peterburg. Otdel. Mat. Inst. Steklov. (POMI) 240, 268–279. (English translation in J. Math. Sci. 96, 3616–3623). MR1691650

[85]     Vershik, A.M., Yor, M. and Tsilevich, N.V. (2001). Thevyt Markov-Krein identity and the quasi-invariance of the gamma process. Zap. Nauchn. Sem. S.-Peterburg. Otdel. Mat. Inst. Steklov. (POMI) 283, 21–36. (English translation in J. Math. Sci. 121, 2303–2310). MR1879060

[86]     Walker, S.G. and Muliere, P. (1997). Beta-Stacy processes andwm2 a generalization of the Pólya-urn scheme. Ann. Statist. 25, 1762–1780. MR1463574

[87]     Yamato, H. (1980). On behaviors of means of distributions withyamato1 Dirichlet processes. Rep. Fac. Sci. Kagoshima Univ. 13, 41–45. MR0618622

[88]     Yamato, H. (1984a). Expectations of functions of samples fromyamato2 distributions chosen from Dirichlet processes. Rep. Fac. Sci. Kagoshima Univ. Math. Phys. Chem. 17, 1–8. MR0798590

[89]     Yamato, H. (1984b). Characteristic functions of means ofyamato distributions chosen from a Dirichlet process. Ann. Probab. 12, 262–267. MR0723745

[90]     Yano, K. and Yano, Y. (2008). Remarks on the density of the law ofyanos the occupation time for Bessel bridges and stable excursions. Statist. Prob. Letters 78, 2175-2180. MR2458027




Home | Current | Past volumes | About | Login | Notify | Contact | Search

Statistics Surveys. ISSN: 1935-7516