Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Exponential distribution as a stress-strength model with type-I censored data from a bayesian viewpoint Abd Elfattah, A. M. Department of Mathematical Statistics, Institute of Statistical Studies &Research, Cairo University, Cairo, Egypt E-Mail: [email protected] Marwa O. Mohamed Department of Mathematics, Zagazig University, Cairo, Egypt E-Mail: [email protected] Abstract-In this paper, the Bayes estimate is derived for the parameters of the exponential model. The estimate is obtained using the squared error loss and LINEX loss function. The risk with the estimate of , under LINEX loss function has been made. Finally, numerical study is given to illustrate the results. Keywords: Exponential distribution, Bayes’ estimator, LINEX loss function, Reliability function, stress-strength model, Risk function, Squared error loss function. Acronyms and Abbreviations CDF PDF MLE UMVUE MSE Cumulative distribution function Probability density function Maximum likelihood estimator Uniformly minimum variance unbiased estimator Mean square error Notations F1(:) F2(:) h (S ) Cumulative distribution function of X Cumulative distribution function of Y Convex loss function Density function of S E[ L()] Expected value of L (; ) R L (ˆ , ˆ ) Joint Risk Efficiency of ̂ and ˆ 1. INTRUDUCTION Let X be the strength of the component which is subjected to Stress Y where X and Y are random variables distributed as exponential distribution with parameter and , respectively. The probability density functions of X and Y is given by: x 1 f ( x; ) e , x 0, 0 (1) y 1 f ( y; ) e , y 0, 0 (2) And their cumulative distribution function x (3) F ( x; ) (1 e ). 1 F ( y; ) (1 e y ). (4) Very wide application of exponential distribution in life testing problems. Where, many authors have studied the stress-strength model with exponential distribution. For example, Tong (1977) had a look at the estimation of P(Y X) for exponential families. Beg (1980a,b and c) estimated the exponential family, two parameter exponential distribution and truncation parameter distributions. Sathe and Shah (1981) studied estimation of Pr(Y X) for the exponential distribution. Awad and Charraf (1986) studied three different estimators for reliability has a bivariate exponential distribution. Kunchur and Mounoli (1993) obtained UMVUE of stressstrength model for multi component survival model based on exponential distribution for parallel system. Shrinkage estimation of Pr(Y X) in the exponential case discussed by Ayman and Walid (2003). In the estimation of reliability function, use of symmetric loss function may be inappropriate as has been recognized by Canfield (1970) and Varian (1975). Zellner (1986) proposed an asymmetric loss function known as the LINEX loss function, which has been found to be appropriate in the situation where overestimation is more serious Under-estimation or Vice-versa. Also, in recent time there are a lot of authors studied LINEX loss functions for different distribution. Such as, Jaheen (2004) studied the exponential model based on record statistics, where the Bayes estimator using the square loss function and LINEX loss function. Bayes estimator of Exponential distribution studied again in (2006) by Cuirong, Dongchu and Dipak to find the risk function. Also, Sanku (2007) derived Bayes estimators for inverted exponential distribution and obtained squared error and LINEX loss functions. Based on upper record values, the Bayes estimators for the parameter of Rayleigh distribution are obtained. These estimators are obtained On the basis of square error and LINEX (linear-exponential) loss functions was studied by Hendi, Abu-Youssef and Alraddadi (2007). ˆ 1 , where ̂ is an estimate of . Consider the following convex loss function. L(1 ) e c c11 1; c1 0 (5) The sign and magnitude of ‘ c1 ’ represent, respectively, the direction and degree of asymmetry. A positive value of ‘ c1 ’is used when overestimation is more costly than underestimation; while a negative value of ‘ c1 ’ is used in the reverse situation. For ‘ c1 ’ close to zero, this loss function is Suppose 1 1 1 approximately squared error loss and therefore almost symmetric. Several authors (Basu & Ebrahimi, 1991; Rojo, 1987; Soliman, 2000; Zellner, 1986) have used this loss function in various estimation and prediction problems. Also, the same definition for where ˆ 2 1 and ˆ is an estimate . Consider the following convex loss function. L( 2 ) ec2 2 c2 2 1; c2 0 The sign of ‘ c2 ’is treated like ‘ c1 ’. 2. BAYES’ ESTIMATE OF and 2 (6) In this section we are concerned with the estimation of the unknown parameters and of the exponential distribution based on a type-I censored random sample of size n and m , respectively. The likelihood function of is given by: L( x1 ,..., xn | ) ( and for 1 n xi i T1 ( n U ) i 1 1 L( y1 ,..., ym | ) ( )V e m i 1 j 1 (7) m y j j T2 ( mV ) j 1 n )U e (8) Put S1 xi i T1 (n U ) and S 2 y j j T2 ( m V ) Because xi and yj ; i 1,..., n and j 1,..., m are independent and identically distributed exponential random variables with Parameters and ,respectively. It is also known that a sum of independent exponentially distributed random variables gives a Gamma distributed variable where the probability density functions of S1 and S 2 is S1 1 1 n1 n i h( S1 ) ( ) S1 e , S1 0 (n 1) i (9) and S2 1 1 m1 m j h( S 2 ) ( ) S2 e , S2 0 (m 1) j It follows, from (7) and (9), that the posterior density of , for a given π 1 e n ( xi i b1 T1 ( n U )) i 1 α (10) x , is given by n (b1 xi i T1 ( n U )) a1 U i 1 a1 U 1 ( a1 U ) (11) Under the square error loss function, the Bayes estimator of , denoted by ˆ MSB1 , is the mean of posterior distribution that can be shown to be ˆ MSB1 b1 S1 a1 U 1 Also, from (8) and (10), the posterior density of , for a given e π 2 m ( y j j b2 T2 ( m V )) j 1 α (12) y , is given by m (b2 y j j T2 ( m V )) a2 V j 1 a2 V 1 ( a2 V ) (13) Under the square error loss function, the Bayes estimator of , denoted by ˆMSB2 , is the mean of posterior distribution that can be shown to be 3 ˆ MSB2 b2 S 2 a2 V 1 (14) 2.1. Bayes’ Estimator of and , based on LINEX Loss Function Under the LINEX loss function (5), the posterior expectation of the loss function L(1 ) with respect to π ( | x) in (11) for the case of 1 E[ L(1 )] (e c1 ( ˆ 1) c1 ( 0 c1 e E (e c1 ( ˆ ) ˆ 1) 1)π ( | x)d 1 ˆ ) c1 E ( ) 1 (15) The value of ̂ that minimizes the posterior expectation of the loss function L(1 ) denoted by ˆ LB1 is obtained by solving equation ˆ E[ L(1 )] c c1 ( ) 1 e c1 E ( 1 e ) c1 E ( ) 0 ˆ that is, ˆ LB1 is the solution of the equation E( 1 e c1 ( ˆ ) 1 ) ec1 E ( ) (16) (17) provided that all expectation exists and are finite. Using (11) and (17), we get the optimal estimate of ̂ relative to L(1 ) c1 ˆ LB1 (b S1 ) 1 (1 e a1 U 1 ) c1 (18) Similarly, the LINEX loss function (6), the posterior expectation of the loss function L( 2 ) with respect to π ( | y) in (12) for the case of , the value of ˆ that minimizes the posterior 2 expectation of the loss function L( 2 ) denoted by ˆ LB 2 is c2 (b S ) ˆLB 2 2 2 (1 e a2 V 1 ) (19) c2 2.2. the joint Risk Efficiency of ˆ LB1 and ˆ LB 2 with respect to ˆ MSB1 and ˆMSB2 under LINEX Loss L(1 ) and L( 2 ) ˆ LB1 and ˆLB 2 relative to L(1 ) and L( 2 ) are of interest. These joint risk functions are denoted by RL (ˆ LB1 , ˆLB 2 ) and RL (ˆ MSB1 , ˆMSB2 ) , where subscript L The risk function of estimators denotes risk relative to L(1 ) and L( 2 ) and are given by using h( S1 ) and h(S2 ) . Let us study the joint risk efficiency of ˆ LB1 and ˆLB 2 , we will find RL (ˆ LB1 , ˆLB 2 ) as follow 4 RL (ˆ LB1 , ˆ LB 2 ) E XY ( L ( 1 ), L ( 2 )) L ( 1 ) L ( 2 ) h ( S1 ) h ( S 2 ) dS1d S 2 00 ˆ c ( LB1 1) ˆ [e 1 c1 ( LB1 1) 1]h ( S1 ) dS1 0 ˆLB 2 c ( 1) 2 ˆ LB 2 c2 ( 1) 1]h ( S 2 ) dS 2 [e 0 c1 a1 U 1 b1 (1 e ) R L(αˆ LB1,βˆ LB 2 ) [ e c1 c1 a U 1 (1 (1 e 1 )i ) n 1 b2( 1 e [ e c2 a2 V 1 β (1 e c1 a1 U 1 )( b1 (20) (n 1)) c1 1] (21) ) c2 c2 a V 1 (1 (1 e 2 )γ j ) m 1 (1 e c2 a2 V 1 )( b2 (m 1)) c2 1] In the same manner, we get RL (ˆ MSB1 , ˆ MSB 2 ) E XY ( L ( 1 ), L ( 2 )) L ( 1 ) L ( 2 ) h ( S1 ) h ( S 2 ) dS1d S 2 00 ˆ c ( MSB1 1) ˆ [e 1 c1 ( MSB1 1) 1]h ( S1 ) dS1 ˆ 0 ˆMSB 2 c ( 1) 2 ˆ MSB 2 c2 ( 1) 1]h ( S 2 ) dS 2 [e 0 5 (22) c1 ( b1 ( a1 U 1) (a U 1) i e R L (ˆ MSE1 , ˆ MSE 2 ) [ 1 (c1i (a1 U 1)) n 1 n 1 c 1 1 (b1 ( n 1)) c1 ( a U 1) 1 m 1 c2 ( b2 ( a2 V 1) (a V 1) e [ 2 (c 2 j (a 2 V 1)) m 1 1) 1] (23) 1) c (a 2 1 2 (b2 ( m 1)) c 2 V 1) The risk efficiency of RL (ˆ LB1 , ˆLB 2 ) with respect to L(1 ) and L( 2 ) may be defined as follows: RE1L (ˆ LB1 , ˆ LB 2 , ˆ MSB1 1] RL (ˆ MSB1 , ˆMSB2 ) under LINEX Loss RL (ˆ MSB1 , ˆMSB2 ) MSB2 ) RL (ˆ LB1 , ˆLB 2 ) , ˆ (24) 3. NUMERICAL Examples To compare the proposed estimator ˆ LB1 and ˆ LB 2 with the estimator ˆ MSB1 and ˆMSB2 , the risk functions are computed so as to see whether RL (ˆ LB1 , ˆLB 2 ) out performs RL (ˆ MSB1 , ˆMSB2 ) under LINEX loss L(1 ) and L(2 ) . A comparison of this type may be needed to check whether an estimator is inadmissible under some loss function. If so, the estimator would not be used for the losses specified by that loss function. For this purpose the risks of the estimators and risk efficiency have been computed RE1L . Since one sample does not tell us much, so we generated N=1000 samples of sizes n, m 5,10,15 and 20 and The results are presented in tables (1-8) where and 50. It is evident from that the values of RE1L c1 c2 0.8,1 2 and 2,3 . and -1 and our time is T=30 is decreasing when n, m increase. Expect for Table n m the values of RE1L is increasing as n, m increasing. But, for n m the values of RE1L is decreasing as n, m increasing. Also, in all the risk efficiency RE1L is greater than one for the sample sizes n, m 5,10,15 and 20 and for all values of c1 , c2 , which indicates that the proposed estimators RL (ˆ LB1 , ˆ LB 2 ) is preferable to RL (ˆ MSB1 , ˆ MSB2 ) . (1) at u1=0.8,u2=0.8,α =2,β=3 has another values because when 3. Acknowledgment Deeply thanks to the referees from the authors, deeply grateful to associate editor and the editor of the Journal their extremely helpful comments and valued suggestions that led to this improved version of the paper. 6 4. REFRENCES Basu, A., Ebrahimi, N. (1991) Bayesian Approach to life Testing and Reliability Estimation Using Asymmetric Loss Function. Jour. Stat. Plann. Infer. 29, 21-31. Canfield, R. (1970) A Bayesian Approach to Reliability Estimation Using a Loss Function. IEEE Transaction on Reliability R-19, 13-16. Cuirong, R., Dongchu, S. and Dipak, K. D.(2006). Bayesian and frequentist estimation and prediction for exponential distribution. Journal of Statistical Planning an inferences, (136), 28732897. Hendi, M. I., Abu-Youssef, S. E. and Alraddadi, A. A.(2007). A Bayesian Analysis of Record Statistics. International Mathematical Forum, 2, no. 13, 619 – 631. Jaheen, Z. F. (2004). Empirical Bayes analysis of record statistics based on LINEX and Quadratic loss function. Computer and Mathematics with Applications, (47), 947-954. Rojo.J. (1987) On the admissibility of CX+d with respect to the LINEX Loss Function. Commun. Statis.-Theory.Meth, 16, 3745-3748. Sanku D. (2007). Inverted exponential distribution as a life distribution model from a bayesian viewpoint. Data Science Journal, Volume 6,107-113. Soliman, A. (2000) Comparison of Linex and Quadratic Bayes estimators for the Rayleigh distribution. Commun. Statist.-Theory Meth. 29(1), 95-107. Varian, H.R. (1975) A Bayesian Approach to Reliability Real Estate Assessment. Amsterdam, North Holland, 195-208. Zellner, A. (1986) Bayesian Estimation and Prediction Using Asymmetric Loss Functions. Jour. Amer. Statist.Assoc.81, 446-451.Data Science Journals, Volume 6. 7