- Open Access
The maximum likelihood estimations for a type of left ellipsoidal distribution
Journal of Inequalities and Applications volume 2013, Article number: 502 (2013)
In this paper, we prove that if the distribution of an random matrix Y is a left ellipsoidal distribution with parameter , , and are independent and identical distributions, the maximum likelihood estimations of μ, Σ are , if and only if . If are not independent and identical distributions, then Y may be not a normal distribution.
Let , be independent standard normal distributions. A random matrix is said to be a left stochastic ellipsoid matrix if it satisfies (see )
and its distribution, which is denoted by , is called a left ellipsoidal distribution, where is reversible, and E is ellipsoid. Besides, d denotes identical distribution, S denotes , and all are orthogonal matrices with order n. Then we have conclusions:
where k is an arbitrary integer and k is a constant (see ).
If X has the distribution density, the distribution density has the form , then the distribution density is said to be spherical distribution. If the distribution density of Y is
then the distribution density is called left ellipsoidal distribution contour, and the contour surface is
When , it is an ellipsoidal surface. The spectral decomposition of Σ is , where Γ is an orthogonal matrix, , .
If the distribution density of Y is
the elongated vector of Y, which is denoted by , is normal distribution. For the sake of consistency of symbols, we denote the normal distribution by .
Classical statistical analysis is built on the basis of normal distribution. However, it is still an important problem whether these graceful properties can also be satisfied without the condition that it is a normal distribution.
Since the left ellipsoidal distribution family has much more members than the normal distribution family (in fact, it almost includes all common distributions), a large amount of scholars’ fruitful research shows: on the one hand, the left ellipsoidal distribution as a multivariate normal distribution promotion is ideal; on the other hand, based on the research of the left ellipsoidal distribution, we can get many statistics used as solid properties of hypothesis test (see ). It is a trivial idea to extend the properties of the normal distribution family to the left ellipsoidal distribution family. Nevertheless, that is not always true. In this article, we prove that the maximum likelihood property of normal distribution cannot be extended to the left ellipsoidal distribution.
Let be the samples of independent identical distributions, and in the condition of normal distribution,
are the maximum likelihood estimations of μ, Σ, respectively, if m is big enough and is positive definite. When , is positive definite with probability 1 if (see ). Additionally, the rank of does not decrease with the increasing of the columns in Y so that is positive definite when . Now we discuss them in the case of . When one focuses on the maximum likelihood estimation, the likelihood equations need to be deduced by the matrix differential method. Therefore, we add a trivial condition that the distribution density of is differentiable (see [5–8]).
We draw the following conclusions.
Theorem 1.1 Let be independent identically distributed, and . The maximum likelihood estimation of μ, Σ is , if and only if .
Note that for the proof of Theorem 1.1 one can refer to .
When the variables of are not independent, we have the solution as follows.
Theorem 1.2 Let , , and . Besides, , where is an matrix. Then if are mutually independent identically distributed, the maximum likelihood estimations of μ, Σ are
if and only if , where , we note it in the proof below.
Theorem 1.3 Under the conditions of Theorem 1.2, if are not independent identically distributed, and the maximum likelihood estimations of μ, Σ are
then Y may be not a normal distribution.
2 The proof of Theorem 1.2
Let be mutually identically independent distributions, and . Besides, we assume that , and , then
where Γ is a random orthogonal matrix with order n. As a result, .
Now the above proposition can be transformed to a research of the form of the left stochastic ellipsoid distribution with a degraded mean.
where , . Therefore, utilizing the single mean degraded left stochastic ellipsoid to express Theorem 1.1, we can obtain Theorem 1.2.
3 The proof of Theorem 1.3
Based on the above discussion, we get the question: If the condition of mutually identically independent distribution is discarded, can one obtain the same solution with Theorem 1.1 from ? In other words, can the proposition be proved or not?
Let , and then the maximum likelihood estimations of μ, Σ are
Here, if and only if .
The proposition is proved not to be true through the above discussion, and we can give the paradoxical instance. Now, we firstly deduce the necessary and sufficient conditions that the maximum likelihood estimation of the parameters μ, Σ in the left stochastic ellipsoid is , .
We assume that Y has the distribution density, namely the likelihood function is
Computing the logarithm difference of both sides, we can get
where and .
Let , and then the maximum likelihood estimation of μ, Σ, which is denoted by , , satisfies
Then we can obtain
As a result, we can conclude that
where C is a random matrix.
Since , . Besides, the 2-order differential of a likelihood function needs to satisfy: .
Consequently, we have the conclusion as follows.
Let , and then the maximum likelihood estimations of μ, Σ are , if and only if
Since if and only if , Y may not be a normal random matrix in general.
For instance, let , , , and .
Then the distribution density of Y is
Since and , and so that the maximum likelihood estimations of μ, Σ are , , then the distribution of Y is not a normal distribution.
In this paper, we proved that if the distribution of an random matrix Y is a left ellipsoidal distribution with parameter , , and are independent and identical distributions, and , then the maximum likelihood estimations of μ, Σ are , . If are not independent and identical distributions, then the maximum likelihood estimations of μ, Σ are , . We used the matrix differential method to deduce that only and only if , which needs to satisfy two conditions.
Zhang, Y: Some Distribution Results for Spherical Distribution and Their Applications, 405. Econometrics and Statistics Colloquium, Rosenwald (1980)
Wang L: On the moments of a kind of elliptical matrix distributions. Chinese J. Appl. Probab. Statist. 1987, 36(3):231–242.
Zhang Y, Fang K, Chen H: The matrix distribution family regression. J. Math. Phys. 1985, 5(3):341–353.
Ahsanullan M: Some characterizations of the bivariate normal distribution. Metrika 1985, 32: 215–218. 10.1007/BF01897814
Nguyen, TT: A note on matrix variate normal distribution. Technical report, Dept. of Math. and Statist., Bowling Green State University, Bowling Green, Ohio (1993)
Zhang, Y, Fang, K: In Multivariate Statistical Analysis, 81. Science Press (1982)
Zhang Y: Matrix differential - the means to deducing exact distribution. J. Math. Res. Expo. 1983, 3(1):179–184.
Anderson, TW, Fang, KT: Maximum likelihood estimators and likelihood ratio criteria for multivariate elliptically contoured distributions. Technical report No. 1, ARO Contract DAAG 29–82-K-0156, Department of Statistics, Stanford University (1982)
Wang Y, Zhang Y: The maximum likelihood estimations for left ellipsoidal distribution. J. Henan Norm. Univ. Nat. Sci. 2008, 36(4):17–19.
This work is supported by the Foundation and Frontier Technology Research Project of Henan Province of China (102300410264) and the Education Department Fundamental Research Project of Henan Province of China (2010A110010). The authors would like to thank the anonymous referees for several useful interesting comments and suggestions about the paper.
The authors declare that they have no competing interests.
All authors contributed equally to the manuscript, read and approved the final manuscript.
About this article
Cite this article
Wang, Y., Zhang, Y. The maximum likelihood estimations for a type of left ellipsoidal distribution. J Inequal Appl 2013, 502 (2013). https://doi.org/10.1186/1029-242X-2013-502
- left ellipsoidal distribution
- normal distribution
- maximum likelihood estimation