Home| Journals | Statistics Online Expert | About Us | Contact Us
    About this Journal  | Table of Contents
Untitled Document

[Abstract] [PDF] [HTML] [Linked References]

Exact Moments of Lower Generalized Order Statistics from Exponentiated Weibull Distribution and Its Characterization

 

R.U. Khan*, Anamika Kulshrestha and D. Kumar

Department of Statistics and Operations Research, Aligarh Muslim University, Aligarh-202 002, Uttar Pradesh, INDIA.

*Corresponding Address:

[email protected]

Research Article

 

Abstract: The concept of lower generalized order statistics was given by Pawlas and Szynal (2001). Later Burkschat et al. (2003) introduced it as dual generalized order statistics to enables a common approach to descending ordered random variables like reversed ordered order statistics, lower 1 records and lower Pfeifer records. With this definition we shall give explicit expressions for single and product moments of lower generalized order statistics from exponentiated Weibull distribution. Further, some of its important deductions and particular cases are discussed and a characterizing result of exponentiated Weibull distribution have been obtained using conditional moments of lower generalized order statistics.

Key words: Lower generalized order statistics, order statistics, lower record values, single and product moments, exponentiated Weibull distribution and characterization.

2010 Mathematics Subject Classification:   62G30, 62E10

 

  1. Introduction

Kamps (1995) introduced the concept of generalized order statistics2. It is know that order statistics, upper record values and sequential order statistics are special cases of3. In this paper we will consider the lower generalized order statistics 4 which is given as

Let 5, 6, 7, be the parameters such that

8, for all 9.

Then 1 are called 2 from an absolutely continuous distribution function 34 with probability density function 56 if their joint 7 has the form

        0        (1.1)

for   1

The marginal 8 of 2th 3, 4, is

        6.       (1.2)

The joint 9 of 5 and 7, 8, is expressed from (1.1) as

        9

                                                                                                                                1,   2,        (1.3)

where

        1,   1

and

        1,   1.

We shall also take 1. If 1, 1, then 1 reduces to the 1th order statistic, 1 from the sample 1 and when 1, then 1 reduces to the 1th lower 1 record value [Pawlas and Szynal, 2001].

Various developments on lower generalized order statistics and related topics have been studied by Pawlas and Szynal (2001), Ahsanullah (2004, 2005), Mbah and Ahsanullah (2007), Khan et al. (2008), Ather et al. (2010). Khan and Kumar (2010, 2011) and references therein.

In this study, we shall derive the explicit expressions for single and product moments of lower generalized order statistics from the exponentiated Weibull distribution. Further its various deductions and particular cases are discussed and a characterizing result of this distribution is stated and proved.

A random variable 1 is said to have exponentiated Weibull distribution if its 1 is given by

        1,   1, 1, 1.     (1.4)

and the corresponding 1 is

                1,   1, 1, 1.         (1.5)

For more details on this distribution and its applications one may refer to Mudholkar and Srivastava (1993), Mudholkar et al. (1995), Mudholkar and Hutson (1996), Jiang and Murthy (1999) and Nassar and Eissa (2003).

 

  1. Single Moments

We shall first establish some basic results which may be helpful in proving the main result.

Lemma 2.1:  For the exponentiated Weibull distribution as given in (1.5) and any non-negative finite integers 1 and 1,

        1,    1    (2.1)

                                1,           1,                  (2.2)

where

        1.     (2.3)

Proof:  On expanding 1 binomially in (2.3), we get when 1

        1,       (2.4)

where

        1.

By setting 1 in (2.4), we obtain

        1.

We have [Balakrishnan and Cohen (1991), p - 44]

        11,   1,     (2.5)

where  1 is the coefficient of 1 in the expansion 1.

Therefore,

        1

and hence the result given in (2.1).

When 1 that

        1  as  1.

Since (2.1) is 1 form at 1, therefore, we have

        1,     (2.6)

where

        1.

Differentiating numerator and denominator of (2.6) 1 times with respect to 1, we get

        1.

Upon applying L’ Hospital rule, we have

        1.                (2.7)

But for all integers 1 and for all real numbers 1, we have Ruiz (1996)

        1.       (2.8)

Therefore,

        1.                    (2.9)

Now on substituting (2.9) in (2.7), we have the result given in (2.2).

Theorem 2.1:  For exponentiated Weibull distribution as given in (1.5) and 1, 1, 1,

        1       (2.10)

                                                                  1,    (2.11)

where 1 is as defined in (2.3).

Proof:  From (1.2), we have

        1

and hence the result given in (2.10).

On using Lemma 2.1 in (2.10), we have the result given in (2.11).

Identity 2.1:  For 1,  1,  1  and  1,

        1.           (2.12)   

Proof:  At 1 in (2.11), we have

        1.

Note that, if 1, then

        1,  1  and  1,     1    [see Balakrishnan and Cohen (1991)]

and hence the result given in (2.12).

Remark 2.1: Setting 1 in (2.11), the results for single moments are deduced for generalized exponential distribution, which verify the result of Khan and Kumar (2011).

Remark 2.2:  Putting  1, 1 in (2.11), the explicit expression for the single moments of  order statistics of the exponentiated Weibull distribution can be obtained as

        1.

That is

        1,

where                   

        1.

Remark 2.3:  Putting  1 in (2.11), we deduce the explicit expression for the single moment of lower 1 record values for  exponentiated Weibull distribution in view of (2.10) and (2.2) in the form

        1

                                                                  1

and hence for lower records

        11.

  1. Product Moments

Before coming to the main results we shall prove the following Lemmas.

Lemma 3.1:  For the distribution as given in (1.5) and non-negative integers 1, 1and 1 with 1,

        1

                                                1,                   (3.1)

where

        1.       (3.2)

Proof:  From (3.2), we have

        1,       (3.3)

where

        1.      (3.4)

Making the substitution 1 in (3.4), we get

        1.      (3.5)

On using (2.5) in (3.5) and simplifying the resulting expression, we obtain

        1.

On substituting the above expression of 1 in (3.3), we find that

        1.    (3.6)

Again by setting 1 in (3.6) and then using (2.5) in resulting expression, we get the relation given in (3.1).

Lemma 3.2:  For the distribution as given in (1.5) and any non-negative integers 1, 1and 1,

        1

1,   1        (3.7)

1

                                                1,                   1,       (3.8)

where 1 is as given in (3.2).

Proof:  When 1, we have

         1

                                                                                                1.

Now substituting for 1 in (3.2), we get when 1

        1

                                    1.

Making use of the Lemma 3.1, we derive the relation given in (3.7).

When 1, we have

        1,  as  1.

On applying L’ Hospital rule and then using (2.9), (3.8) can be proved on the lines of (2.2).

Theorem 3.1:  For exponentiated Weibull distribution as given in (1.5) and for 1, 1, 1,

        1

                                                                                                                                1        (3.9)

                                                                                                                    1

                                                                                                1.    (3.10)

Proof:  From (1.3), we have

1

                                1

                                1 .               (3.11)

On expanding  1 binomially in (3.11), we get

        1

                                                                                                1

                                                1

and hence the result given in (3.9).

Now on making use of the lemma 3.2 in (3.9), we derive the relation given in (3.10).

Identity 3.1:  For 1, 1,  1,  1 and 1,

        1.      (3.12)

Proof:    At 1 in (3.10), we have

        1

                1.

Note that, if  1, then

        1,  1,  1  and

        1,  1,  1.  [see Balakrishnan and Cohen (1991)]

Therefore,

        1.

Now on using (2.12), we get the result given in (3.12).

At 1, (3.12) reduces to (2.12).

Remark 3.1:  Putting 1 in (3.10), the results for product moments of Khan and Kumar (2.11) for generalized exponential distribution are deduced.

Remark 3.2:  Setting 1, 1 in (3.10), the explicit expression for the product moments of  order statistics of the exponentiated Weibull distribution can be obtained as

        1

                                                1.

That is

        1

                                                                1,

where

        1.

Remark 3.3:  Putting 1 in (3.10), we deduce the explicit expression for the product moments of lower 1 record values for the exponentiated Weibull distribution in view of (3.9) and (3.8) in the form

        1

and hence for lower records

        1.

Remark 3.4:  At 1 in (3.10), we have

        1

                                                                1,

where

        1,  1  and  1,  1.

Therefore,

        1

                                                                                1.             (3.13)

Making use of (3.12) in (3.13) and simplifying the resulting expression, we get

        1

                                                                                1

as obtained in (2.11).

 

  1. Characterization

Let 1, 1 be 1 from a continuous population with 1 1 and  1  1, then the conditional 1 of 1 given 1, 1, in view of (1.3) and (1.2) is

                1

                                                                                1.                       (4.1)

Theorem 4.1:  Let 1 be a non negative random variable having an absolutely continuous distribution function 1 with 1 and 1 for all 1, then

        1

                                                                                                1,     1                            (4.2)

if and only if

        1,  1,  1,  1,

where  1.

Proof:  From (4.1), we have

        1

                                                1.                 (4.3)

By setting 1 from (1.5) in (4.3), we find that

        1

                                                                                                                                                                1

                1,          (4.4)

where

        1.

Again by setting  1 in (4.4), we get

        1

                                                1

                1

                1.

Simplifying the resulting expression, we derive the relation in (4.2).

To prove sufficient part, we have from (4.1) and (4.2)

        1

                                                                1,           (4.5)

where

        1.

Differentiating (4.5) both sides with respect to 1, we get

        1

                                1

or

        1,

where

        1

                                                1

and

        1

                                                                                1.

Therefore,

        1

                   1

which proves that

        1,   1, 11, 1.

Acknowledgments

The authors are thankful to the referee and the editor for their valuable suggestions to improve the quality of this paper.

 

References

    1. Ahsanullah, M. (2004): A characterization of the uniform distribution by dual generalized order statistics. Comm. Statist. Theory Methods, 33, 2921-2928.
    2. Ahsanullah, M. (2005):  On lower generalized order statistics and a characterization of power function distribution. Stat. Methods, 7, 16-28.
    3. Athar, H., Khan, R.U. and Anwar, Z. (2010): Exact moments of lower generalized order statistics from power function distribution. Calcutta Statist. Assoc. Bull., 62, 245-246.
    4. Balakrishnan, N. and Cohen, A.C. (1991):  Order Statistics and Inference: Estimation Methods. Academic Press, San Diego.
    5. Burkschat, M., Cramer, E. and Kamps, U. (2003):  Dual generalized order statistics. Metron, LXI, 13-26.
    6. Jiang, R. and Murthy, O.W.P. (1999):  The exponentiated Weibull family: a graphical approach, IEEE Transaction Reliability, 48, 68-72.
    7. Kamps, U. (1995):  A concept of generalized order statistics.  B.G. Teubner Stuttgart.
    8. Khan, R.U. and Kumar, D. (2010): On moments of generalized order statistics from exponentiated Pareto distribution and its characterization. Appl. Math. Sci. (Ruse), 4, 2711-2722.
    9. Khan, R.U. and Kumar, D. (2011): Expectation identities of lower generalized order statistics from generalized exponential distribution and a characterization. Math. Methods Statist., 20, 150-157.
    10. Khan, R.U., Anwar, Z. and Athar, H. (2008):  Recurrence relations for single and product moments of generalized order statistics from exponentiated Weibull distribution. Aligarh J. Statist., 28, 37-45.
    11. Mbah, A.K. and Ahsanullah, M. (2007):  Some characterization of the power function distribution based on lower generalized order statistics. Pakistan J. Statist., 23, 139-146.
    12. Mudholkar, G.S. and Hutson, A.D. (1996): The exponentiated Weibull family: some properties and a flood data application. Comm. Statist. Theory Methods, 25, 3059-3083.
    13. Mudholkar, G.S. and Srivastava, D.K. (1993): Exponentiated Weibull family for analyzing bathtub failure rate data, IEEE Transactions on Reliability, 42, 299-302.
    14. Mudholkar, G.S., Srivastava, D.K. and Freimer, M. (1995): The exponentiated Weibull family. Technometrics, 37, 436-445.
    15. Nassar, M.M. and Eissa, F.H. (2003): On exponentiated Weibull distribution. Comm. Statist. Theory Methods, 32, 1317-1336.
    16. Pawlas, P. and Szynal, D. (2001): Recurrence relations for single and product moments of lower generalized order statistics from the inverse Weibull distribution. Demonstratio Math., XXXIV, 353-358.
    17. Ruiz, S. M. (1996):  An algebraic identity leading to Wilson’s theorem.  Math. Gaz., 80, 579-582.

     
 
 
 
 
 
  Copyrights statperson consultancy www

Copyrights � statperson consultancy www.statperson.com  2013. All Rights Reserved.

Developer Details