1973
DOI: 10.1214/aop/1176996898
|View full text |Cite
|
Sign up to set email alerts
|

Contributions to the Theory of Dirichlet Processes

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
73
0
1

Year Published

1997
1997
2019
2019

Publication Types

Select...
6
3

Relationship

0
9

Authors

Journals

citations
Cited by 163 publications
(74 citation statements)
references
References 0 publications
0
73
0
1
Order By: Relevance
“…1 with probability 1 as n ! 1 (Korwar & Hollander, 1973), the number of components increases approximately logarithmically with the number of observations. This is illustrated in Fig.…”
Section: Learning the Dispersion Of Datamentioning
confidence: 94%
“…1 with probability 1 as n ! 1 (Korwar & Hollander, 1973), the number of components increases approximately logarithmically with the number of observations. This is illustrated in Fig.…”
Section: Learning the Dispersion Of Datamentioning
confidence: 94%
“…From the form of the predictive distribution, proceeding as in Korwar and Hollander (1973) and Antoniak (1974, Lemma 1 ), it can be shown that P, i.e. the joint probability law of (X1 ..... X,), gives positive probability to certain collections of hyperplanes in the sample space R ~.…”
Section: The Likelihood Of a Sample From A Diriehlet Processmentioning
confidence: 99%
“…the joint probability law of (X1 ..... X,), gives positive probability to certain collections of hyperplanes in the sample space R ~. Let us introduce the set Cg (D, nl ..... riD I~(D'nl,'",no) Korwar and Hollander (1973) show that, given the number D of distinct values, these are i.i.d. according to fo.…”
Section: The Likelihood Of a Sample From A Diriehlet Processmentioning
confidence: 99%
“…Another distinctive feature, if compared with the Dirichlet process, is represented by the asymptotic behaviour of the number of groups K n generated by the first n observations, as n → ∞. For the Dirichlet process, as shown in Korwar and Hollander (1973), K n ∼ θ log(n) almost surely as n → ∞. Hence, the number of distinct observations increases at a logarithmic rate.…”
Section: Example 9 (The Two Parameter Poisson-dirichlet Process) Onmentioning
confidence: 99%