Preprint
Article

On Estimation for Reliability of Stress-Strength Model Based on Topp-Leone Distribution

Altmetrics

Downloads

212

Views

45

Comments

0

This version is not peer-reviewed

Submitted:

08 October 2023

Posted:

09 October 2023

Read the latest preprint version here

Alerts
Abstract
The Paper points at an important point, which is estimate the reliability of system used for stress-strength model when the stress and strength follows Topp-Leone distribution. In this context, this work including two models of stress –strength, the first one when the system has one component with strength subject to one stress, while the another model concern with system has one component which has strength subject for two bounded stresses .The expressions of system reliability of two considered models were derived and estimate using different methods. Comparisons between the considered estimators were made depending on simulation technique based on statistical criterion namely mean squared error.
Keywords: 
Subject: Computer Science and Mathematics  -   Probability and Statistics
Preprints 87244 i001

I. Introduction

One of the most popular techniques for data analysis is the stress-strength system, which is employed in a wide range of disciplines like developed engineering, military presentations, healthiness, and useful skills.
The reliability of system in stress-strength model is an assessment of a module's dependability in terms of the random variable X, which stands in for the stress the module is exposed to, and Y, which stands in for the component's capacity to withstand the potential stress.
In stress–strength model of system, the strength X exposed the stress Y. Both random variables X and Y supposed to follows specific probability distribution with definite parameters. The reliability in stress–strength model refers to the probability that strength overdoes stress, i.e. P(X > Y), roughly p. This topic consumes numerous presentations in several ranges. For example, if Y denotes the extreme heaviness produced by overflowing and X signifies the strength of the leg of a bridge on a stream, then p is the probability that the bridge will be hard. Additional instance, if Y and X are respectively symbolize the regulator and conduct groups, then P processes the treatment consequence. Then the estimation of P will be significant in creation interpretations. The system fails when the stress is too great for it to handle.
The stress-strength concept is particularly significant in the research on reliability. Most of the concerns in the statistical approach to the stress-strength model are predicated on the premise that the component strengths are randomly and uniformly distributed and are exposed to a single stress [1].
A system experiences a stress Y and strength X when demonstrated in a conventional reliability stress-strength analysis. Both random variables are considered to follows specific distribution with known or unknown parameters. The probability of strength exceeding stress, or P(X>Y), indicates the reliability of the system. This topic has many applications in many fields
The stress-strength models of the types P(Y<X), P(Y<X<Z), where X, Y and Z are independent random variables refers to strength X and two stress Y and Z, and its follows specific distribution. These two models have wide requests in several of engineering subareas, psychology, genetics, medical trials and others. Kotz [2].
Isaam, K. , Taha, A. and Abbas,N. They Estimated P(Y<X) using different estimation methods[3] .Chandra and Owen [4] derived maximum likelihood estimators (MLEs) and consequent uniform minimum variance unbiased estimators (UMVUEs) for R= P(Y<X< Z).
Singh [5] offered the minimum variance unbiased, maximum likelihood and empirical estimators of R= P(Y<X<Z), where X, Y and Z are independent random variables and follows the normal distribution. Dutta and Sriwastav [6] estimated R when X, Y and Z are exponentially distributed. Ivshin [7] studied the MLE and UMVUE of R when X, Y and Z are either uniform or exponential random variables with unknown location parameters. Wang et al. [8] make statistical inference for P(X<Y<Z) via two methods, the nonparametric normal approximation and the jackknife empirical likelihood.
The Topp-Leone (TL) distribution is therefore J-shaped through its support. Percentage data, rates, particle sizes and specific chemical procedure yield data that can be displayed by this distribution. The TL distribution has a finite support, and various data sets in reliability and life testing are showed using finite support distributions. [9].
Particularly, when the reliability is measured as the proportion of the quantity of effective trials to the amount of whole trials, the TL distribution can well be functional. In stress–strength model the distributions have uses in many spaces. For example, if Y denotes the maximum section elongation and X signifies the tensile strength of a piece of some material, then p processes the quality of the material. Another example, if Y refers the radius of the base of a small cup and X represents the radius of the circular depression in the center of a saucer then P represents the probability of holding the cup. Also, a consumer research organization may want to compare sales percentages of two products with a different advertisement policy each.
Topp and Leone [10] presented the TL distribution and display its properties and also showed its applications for some failure data. Nadarajah and Kotz [11] derive some properties of the TL distribution and provided an expression for its characteristic function. Kotz and Van Dorp [12] given a generalized TL distribution to model some economic facts and they also clear a reflected general TL distribution and studied its properties. Ghitany et al. [13] considered the related of the reliability function of this distribution such as the hazard rate; mean residual life, reversed hazard rate, expected inactivity time and their stochastic orderings.
The topic of the Topp-Leone (TL) distribution dealt with this research, and the probability density function (pdf) is
f X = 2 α ( 1 x ) x α 1 ( 2 x ) α 1    
where 0 < x < 1 and 0 < α < 1.
The distribution function of the TL distribution is given by
F X = 0   , x 0   , x α ( 2 x ) α 0 < x < 1 1 x 1
So if u follows uniform distribution, then X = 1 1 U 1 / α has the TL   ( α ) distribution.
Consequently, the hazard rate will be as below
h x = f x 1 F x = 2 α ( 1 x ) 1 ( 1 x ) 2 α 1 1 1 ( 1 x ) 2 α
The rest of this paper is structured as follows. In Section II, the expression of R1= P(X > Y) and R2= P(Y<X< Z) will be derived. Maximum likelihood estimator MLE, the Moment estimator MOM and the Pre-test single stage shrinkage estimator SH of R is obtained in Section III. Monte Carlo Simulation and Numerical Outcomes are laid out in Section IV. Finally, conclusions are presented in Section V.

II. Expression of R1= P(X > Y) and R2= P(Y<X< Z)

  • Derivation of R1 = P(X > Y)
This Section concentrates on estimating the reliability of when X and Y have independent Topp-Leone distributions. Let n be the number of observations distributed according to the Topp-Leone
Now let X TL (α) be independent of Y TL (β). Then
R 1 = P ( X > Y ) = 0 1 F Y x f x x d x         R 1 = 2 α 0 1 x α + β 1 1 x 2 x α + β 1 d x       R 1 = α α + β      
Where α and β are unknown.
  • Derivation of R2 = P(Y<X<Z)
This Section concentrates on estimating the reliability of R2 = P(Y<X<Z) when Z, Y and X have independent Topp-Leone distributions such that X, Y and Z are independent and they are distributed Topp-Leone with scale parameters α , β , γ respectively such that the p.d.f of the strength X is
f x x , α = 2 α 1 x x α 1 2 x α 1             x > 0     ,   α > 0   .  
Consequently, the p.d.f of the stresses Y, Z are given respectively by
f y y , α = 2 β 1 y y β 1 2 y β 1             y > 0     ,   β > 0   .  
f Z z , γ = 2 γ 1 z z γ 1 2 z γ 1             z > 0     ,   γ > 0   .  
The reliability system of this model P(Y<X<Z) given by
R 2 = P ( Y < X < Z ) = 0 1 P Y < X , X < Z f ( x ) d x = 0 1 F Y x F ¯ z x f ( x ) d x = 0 1 F Y x f x 0 1 F Y x F z x f ( x ) d x
= α α + β 0 1 x β ( 2 x ) β x γ ( 2 x ) γ 2 α ( 1 x ) x α 1 ( 2 x ) α 1 d x = α α + β 0 1 2 α x ( α + β + γ ) ( 2 x ) α + β + γ 1 ( 1 x ) x α 1 d x = α α + β α α + β + γ
R 2 = α γ ( α + β ) ( α + β + γ )        
Where ,   β and γ are unknown.

III. Estimation of R1 = P(X > Y) and R2 = P(Y<X<Z)

Maximum Likelihood Estimation of R 1 , R 2
The Maximum likelihood estimator MLE technique is an important and commonly estimator, since its has a good property for estimate which is known as invariant property [14] .
This Section deals with MLE of reliability R 1 =P(X>Y) and R 2 =P(Y < X < Z) when X, Y and Z are independent Topp-Leone distribution with scale parameters ( α , β , γ   ) respectively.
Let x1, x2, … . . xn be a random strength sample of size n with p.d.f. as in (eq.3) then let y1 ,y2, … ym. . and z1, z2, … . . zw be the random samples with p.d.f . as in (eq.3) .The Maximum Likelihood function of the observed sample is:
L α , β , γ , x , y , z = i = 1 n f x i j = 1 m f y j k = 1 w f z k = i = 1 n 2 α 1 x i x i α 1 2 x i α 1 j = 1 m 2 β 1 y j y j β 1 2 y j β 1 k = 1 w 2 γ 1 z k z k γ 1 2 z k γ 1
Taking the logarithm for the above likelihood function eq (6) and the partial derivative for the log-likelihood function with respect to unknown parameters α β   a n d γ , respectively and equating the partial derivative to zero to solve this equation:
L n L x i α = n α + i = 1 n log x i 2 x i = 0
L n L y j β = m β + j = 1 m log y j 2 y j = 0
L n L z k γ = w γ + B = 1 w log z B 2 z B = 0
The results of the above equations give MLEs of the parameters:
α ^ m l = n i = 1 n log x i 2 x i
β ^ m l = m j = 1 m log y j 2 y j
γ ^ m l = w B = 1 w log z B 2 z B
We obtain the MLE of R1 and R2 as
R ( M L ) 1 = α ^ m l α ^ m l + β ^ m l             , R ( M L ) 2 = α ^ m l γ ^ m l ( α ^ m l + β ^ m l ) ( α ^ m l + β ^ m l + γ ^ m l )
Moment Estimation Method of R1 and R2
This section concern with the moment estimator method MOM of R1 and R2.The moment estimators of the unknown parameters α , β   and γ will be obtained by equating the population moments with the corresponding sample moments. The population means of random variables X,Y and Z are as below
    μ x = E x = 1 π 2 Γ α + 1 Γ α + 3 2
  μ y = E y = 1 π 2 Γ β + 1 Γ β + 3 2
μ z = E z = 1 π 2 Γ ( γ + 1 ) Γ ( γ + 3 2 )
Suppose that X= ( x1 , x2 , …xn ) be a random sample of size n and Y=(y1 , y2 …ym) be a random sample of size m and Z=(z1 , z2 …zw) be a random sample of size w follows Topp-Leone distribution with unknown scales parameter   α , β   and γ .
Then the means of the first and sample moments are given by
μ x = x ¯ = 1 n   i = 1 n x i       μ y = y ¯ = 1 m   j = 1 m y i     μ z = z ¯ = 1 w   B = 1 w z i
By equating the samples moments with the corresponding population moments, then μ x = μ x , μ y = μ y , μ z = μ z
x ¯ = 1 π 2 Γ α + 1 Γ α + 3 2 , y ¯ = 1 π 2 Γ β + 1 Γ β + 3 2 , z ¯ = 1 π 2 Γ ( γ + 1 ) Γ ( γ + 3 2 )
The moment estimator of α , β   and γ denoted by α ^ m o   ,     β ^ m o and γ ^ m o can be obtained from (14), respectively as,
α ^ m o   = Γ α ° + 3 2 Γ α ° . 2 π 1 x ¯            
β ^ m o = Γ β ° + 3 2 Γ β ° . 2 π 1 y ¯          
γ ^ m o = Γ γ ° + 3 2 Γ γ ° . 2 π 1 z ¯        
Consequently ,we obtain the Moment estimators of R1 and R2 as
R ( M O ) 1 = α ^ m o α ^ m o + β ^ m o             ,       R ( M O ) 2 = α ^ m o γ ^ m o ( α ^ m o + β ^ m o ) ( α ^ m o + β ^ m o + γ ^ m o )
Pre-test single stage shrinkage estimator (SH) of R1 and R2
Some time may we have a prior information value (point guess) of the parameter to be estimated. If this value is in the neighborhood of the accurate value, the shrinkage procedure is valuable to obtain an improved estimator. Thompson in [15], Isaam, K, Taha, A. and Abbas,N.[16] and others suggested shrunken estimators for different distributions when a prior estimate or guess point is available. They indicated that these estimators perform better in the term of mean squared error when a guess value θ0 close to the true value θ. Pre- test estimator is considered for estimating the parameter θ when a guess point (prior estimate) θ0 is available about θ due the past knowledge or similar cases. From the empirical studies it has been established that the shrinkage estimators performs better than the usual estimator when the guess point is very close to the true value of the parameter. Therefore to make sure whether θ is closed to θ0 or not, we may test H0:θ = θ0 against H1: θ ≠ θ0, so we denote by R to the critical region for above test.
Thompson in 1968 recommended shrinking the usual estimator   θ ^ of θ towards the prior guess point θ0 and suggested the estimator   θ ~ = K θ ^ + ( 1 K θ ^ ) θ °   , where   ( 1 K )   represents the experimenters belief in the guess point θ0. He found the estimator θ ~ which is more efficient than usual estimator   θ ^ if the true value θ is close to θ0 (H0 accepted) but may be less efficient otherwise, therefore to resolve the uncertainty that a guess point value is approximately the true value or not, a pre- test of significance may be employed. So he take the usual estimator θ ^ when θ is far away from θ0 (H0 rejected) after he made the pre- test.
Thus, the pre-test shrunken estimator has the following form ; A.N.Salman [17]
θ ~ = K θ ^ + 1 K θ °                                                               i f θ ^ R   θ ^                                                                                                                     i f   θ ^ R
Where R may be pre- test region for acceptance the null hypothesis H0 as we mentioned above, θ ^ is the usual ML estimator of θ and K is a constant shrinkage weight factor such that
0 ≤ K ≤ 1 .
In this research we may assume the region as follow:
R = ( θ ° θ ) 2 M S E
R = θ ° M S E , θ ° + M S E
In this research, we suppose
Case 1 Suppose that
K n = S i n n n 2 K m = S i n m m 2 K w = S i n w w 2 , 0 K n , K m , K w 1 a n d θ = θ ° .
Where,   θ may be referred to   α   , β   a n d γ   .
Thus, the shrinkage estimator of the scale parameters α , β   and γ of the random variables X,Y, Z that follows Topp-Leone distribution will be as follows:
α ~ s h 1 = K n α ^ + 1 K n α °                                                               i f α ^ R   α ^                                                                                                                     i f   α ^ R
β ~ s h 1 = K m β ^ + 1 K m β °                                                               i f β ^ R   β ^                                                                                                                     i f   β ^ R    
γ ~ s h 1 = K w γ ^ + 1 K w γ °                                                               i f γ ^ R   γ ^                                                                                                                     i f   γ ^ R
Case 2 Suppose that
A n = e n 10         ,   A m = e m 10 , A w = e w 10   , 0 A n , A m , A w 1     a n d   θ = θ °   .
Where ,   θ   may be referred to   α   , β   a n d γ   .
α ~ s h 2 = A n α ^ + 1 A n α °                                                               i f α ^ R   α ^                                                                                                                     i f   α ^ R
β ~ s h 2 = A m β ^ + 1 A m β °                                                               i f β ^ R   β ^                                                                                                                     i f   β ^ R
γ ~ s h 2 = A w γ ^ + 1 A w γ °                                                               i f γ ^ R   γ ^                                                                                                                     i f   γ ^ R
we obtain the Mom of R1 and R2 as
R ( S H ) 1 = α ^ S H 1 α ^ S H 1 + β ^ S H 1             , R ( S H ) 2 = α ^ S H 1 γ ^ S H 1 ( α ^ S H 1 + β ^ S H 1 ) ( α ^ S H 1 + β ^ S H 1 + γ ^ S H 1 )                                                              
Consequently , we obtain
R ( S H 2 ) 1 = α ^ S H 2 α ^ S H 2 + β ^ S H 2             , R ( S H 22 = α ^ S H 2 γ ^ S H 2 ( α ^ S H 2 + β ^ S H 2 ) ( α ^ S H 2 + β ^ S H 2 + γ ^ S H 2 )

IV. Monte Carlo Simulation and Numerical Outcomes

An extensive numerical investigation will be conducted in this section to compare the performance of the various estimators for unlike sample sizes and parameter values for the Topp-Leone distribution. The properties investigated result in mean square errors (MSEs). Matlab 2018 statistical software was used for all computations. A simulation results are conducted to examine and compare the performance of the estimates for shape respecting to the MSE. The best estimator has the smallest value of MSE.
The steps for estimating the parameters, R1 = P(X > Y) and R2 = P(Y< X < Z) can be
summarized as follows:
Generate 10000 random samples from Topp-Leone distribution with the sample sizes;(n ,m)= (25,50,75,100) and the parameter values are selected as α = (1.2,1.5,3) and β =( 3 ,1.5,1.2) for R1 .Also (n, m ,w )= (25,25,25),(50,50,50),(75,75,75), (100,100,100), (50,25,25), (75,25,25),
(100,25,25), (25,50,25), (50,75,50), (75,100,75), (25,100,25), (25,25,50), (50,50,75),
(75,75,100) and parameter values are selected as α = (1.5,1.5,1.5), β =( 2.5 ,1.5,3.5) and γ =( 3.5 ,1.5,2.5) for R2.
One can conclude from the simulation results which is used to determine the best consequence of the proposed estimation methods (ML, MO, SH1, SH2) using different samples for the system reliability R1=p(Y<X) and R2 = P(Y< X < Z) grounded the Topp-Leone distribution (T-L). The simulation results of the proposed estimation methods are demonstrated in annexed tables and distinguish that Pre-test single stage shrinkage estimator (SH1) of system reliability R1 , R2 satisfied the smallest mean squared error in overall; this infers that R ^ S H 1 was the best than the others for two considered models.

V. Conclusion

From above results, it observed that in general the best performance of the consider estimators (ML, MO, SH1, SH2,) under the different sample sizes and for the different Parameters of this study is the pre-test single stage shrinkage estimator (SH1) of system reliability R1 , R2 for two considered models .This important method has proven its efficiency in estimation as prior estimate approach to real value which depends on classical estimation method and prior information (initial estimate) as a linear combination and make pretest region to know how close the initial value from the actual value .

Appendix A

Table A1. Shown estimates of R1, when R1= 0.28571, α = 1.2, β = 3.
Table A1. Shown estimates of R1, when R1= 0.28571, α = 1.2, β = 3.
n m R ^ M L R ^ M O R ^ S H 1 R ^ S H 2
25 25 0.22795 0.76549 0.26949 0.65166
50 0.29000 0.69709 0.28616 0.59951
75 0.25700 0.83259 0.27238 0.71323
100 0.30891 0.70636 0.29865 0.60523
50 25 0.22768 0.72594 0.26577 0.60214
50 0.25496 0.79200 0.28253 0.66948
75 0.31010 0.66498 0.28581 0.59447
100 0.38670 0.58104 0.29034 0.5384
75 25 0.38350 0.58687 0.30097 0.54219
50 0.36616 0.64820 0.29003 0.58296
75 0.27682 0.72916 0.28453 0.62165
100 0.32255 0.65676 0.28536 0.57684
100 25 0.31769 0.69842 0.28650 0.60441
50 0.30583 0.72137 0.28373 0.61462
75 0.34595 0.62115 0.28556 0.55979
100 0.24077 0.78389 0.28439 0.66114
Table A2. MSE of R1, when R1= 0.28571, α = 1.2, β = 3.
Table A2. MSE of R1, when R1= 0.28571, α = 1.2, β = 3.
n m ML MO SH1 SH2 Best
25 25 3.3362e-07 2.3018e-05 2.6327e-08 1.3392e-05 SH
50 1.8372e-09 1.6923e-05 1.9773e-11 9.8467e-06 SH
75 8.2468e-08 2.9907e-05 1.7781e-08 1.8277e-05 SH
100 5.3794e-08 1.7695e-05 1.6745e-08 1.0209e-05 SH
50 25 3.3683e-07 1.938e-05 3.9771e-08 1.0012e-05 SH
50 9.4604e-08 2.5633e-05 1.0121e-09 1.4727e-05 SH
75 5.9453e-08 1.4384e-05 9.8482e-13 9.5327e-06 SH
100 1.0197e-06 8.7219e-06 2.1393e-09 6.3848e-06 SH
75 25 9.5616e-07 9.5616e-07 2.3269e-08 6.5779e-06 SH
50 6.4721e-07 1.314e-05 1.8666e-09 8.8355e-06 SH
75 7.9146e-09 1.9664e-05 1.3949e-10 1.1286e-05 SH
100 1.3571e-07 1.3767e-05 1.2546e-11 8.4754e-06 SH
100 25 1.0226e-07 1.7032e-05 6.1133e-11 1.0157e-05 SH
50 4.0479e-08 1.898e-05 3.926e-10 1.0818e-05 SH
75 3.6288e-07 1.1252e-05 2.3293e-12 7.5116e-06 SH
100 2.0201e-07 2.4818e-05 1.7668e-10 1.4095e-05 SH
Table A3. Shown estimation when R1= 0.5, beta1 = 1.5, beta2= 1.5.
Table A3. Shown estimation when R1= 0.5, beta1 = 1.5, beta2= 1.5.
n m ML MO SH1 SH2
25 25 0.52585 0.44398 0.50746 0.47914
50 0.52351 0.34866 0.50910 0.40764
75 0.53547 0.52974 0.50933 0.52313
100 0.52507 0.54220 0.50871 0.53144
50 25 0.47686 0.56972 0.56972 0.53694
50 0.51562 0.53771 0.50141 0.52235
75 0.50862 0.52169 0.50146 0.5146
100 0.53958 0.40917 0.50059 0.44168
75 25 0.51675 0.46870 0.50091 0.47168
50 0.52847 0.53250 0.50018 0.52059
75 0.53742 0.42200 0.50086 0.45271
100 0.55972 0.43488 0.50177 0.46178
100 25 0.47113 0.61394 0.49444 0.55794
50 0.49079 0.53128 0.50097 0.52011
75 0.52326 0.53541 0.50074 0.53282
100 0.53481 0.43170 0.50023 0.45482
Table A4. MSEs of R1 when α = 1.5, β = 1.5.
Table A4. MSEs of R1 when α = 1.5, β = 1.5.
n m ML LS SH1 SH2 Best
25 25 6.6847e-08 3.138e-07 5.561e-09 4.3506e-08 SH1
50 5.5253e-08 2.2904e-06 8.2734e-09 8.5312e-07 SH1
75 1.2578e-07 8.8426e-08 8.6961e-09 5.3482e-08 SH1
100 6.2857e-08 1.7812e-07 7.5812e-09 9.8874e-08 SH1
50 25 5.3536e-08 4.8611e-07 4.537e-10 1.3645e-07 SH1
50 2.4407e-08 1.4221e-07 1.9898e-10 4.9937e-08 SH1
75 7.4261e-09 4.7039e-08 2.1352e-10 2.1311e-08 SH1
100 1.5667e-07 8.2494e-07 3.4689e-11 3.4007e-07 SH1
75 25 2.8048e-08 9.7946e-08 8.2667e-11 8.0202e-08 SH1
50 8.1083e-08 1.0562e-07 3.4017e-12 4.2379e-08 SH1
75 1.4002e-07 6.0841e-07 7.3392e-11 2.2367e-07 SH1
100 3.5666e-07 4.2403e-07 3.1286e-10 1.4606e-07 SH1
100 25 8.332e-08 1.2983e-06 3.0867e-09 3.3573e-07 SH1
50 8.4884e-09 9.7818e-08 9.4062e-11 4.0457e-08 SH1
75 5.4121e-08 1.2542e-07 5.5441e-11 1.0773e-07 SH1
100 1.2119e-07 4.6645e-07 5.3448e-12 2.0412e-07 SH1
Shown estimation when R= 0.71429, beta1 = 3, beta2= 1.2
n m ML MO SH1 SH2
25 25 0.61334 0.34578 0.69064 0.41593
50 0.61151 0.43041 0.69210 0.46764
75 0.72315 0.23630 0.72105 0.35348
100 0.79167 0.21785 0.73058 0.34541
50 25 0.64462 0.22901 0.69405 0.33525
50 0.73712 0.25847 0.71695 0.37117
75 0.67832 0.28617 0.71318 0.38605
100 0.66978 0.27653 0.71211 0.37392
75 25 0.67920 0.24415 0.71638 0.34904
50 0.74174 0.26702 0.71454 0.38343
75 0.74139 0.24329 0.71603 0.36362
100 0.72348 0.27315 0.71518 0.38607
100 25 0.76171 0.42327 0.71933 0.48892
50 0.68467 0.30166 0.71191 0.39779
75 0.73210 0.24035 0.71535 0.35858
100 0.68185 0.31757 0.71511 0.40896
mse
n m ML MO SH1 SH2 Best
25 25 1.0191e-06 1.358e-05 5.5891e-08 8.9014e-06 SH1
50 1.0562e-06 8.0585e-06 4.9205e-08 6.0834e-06 SH1
75 7.8616e-09 2.2847e-05 4.5744e-09 1.3018e-05 SH1
100 5.9891e-07 2.4645e-05 2.6549e-08 1.3607e-05 SH1
50 25 4.853e-07 2.3549e-05 4.095e-08 1.4367e-05 SH1
50 5.2132e-08 2.0777e-05 7.0912e-10 1.1773e-05 SH1
75 1.2934e-07 1.8328e-05 1.2289e-10 1.0774e-05 SH1
100 1.9804e-07 1.9163e-05 4.7435e-10 1.1585e-05 SH1
75 25 1.2307e-07 2.2103e-05 4.3956e-10 1.334e-05 SH1
50 7.5363e-08 2.0004e-05 6.3547e-12 1.0946e-05 SH1
75 7.349e-08 2.2184e-05 3.0541e-10 1.2297e-05 SH1
100 8.4551e-09 1.946e-05 8.0755e-11 1.0773e-05 SH1
100 25 2.6294e-07 1.6149e-05 1.995e-10 9.6001e-06 SH1
50 8.7697e-08 1.7026e-05 5.6266e-10 1.0017e-05 SH1
75 3.174e-08 2.2462e-05 1.1351e-10 1.2653e-05 SH1
100 1.0524e-07 1.5738e-05 6.7611e-11 9.3224e-06 SH1
R= 0.175,   ζ 1 =1.5, ζ 2 =2.5, ζ 3 =3.5and q=10000
Method
n,m,w
Mle Mom SH1 SH2 Best
(25,25,25) R ^
MSE
0.21921
1.9542e-07
0.12232
2.7753e-07
0.18717
1.4807e-08
0.15602
3.6021e-08
SH1
(50,50,50) R ^
MSE
0.15329
4.7154e-08
0.15246
5.0796e-08
0.17317
3.3671e-10
0.17118
1.4578e-09
SH1
(75,75,75) R ^
MSE
0.16695
6.486e-09
0.15469
4.1266e-08
0.17473
7.4067e-12
0.18071
3.2602e-09
SH1
(100,100,100) R ^
MSE
0.20633
9.8161e-08
0.10786
4.5076e-07
0.17513
1.5626e-12
0.13863
1.3226e-07
SH1
(50,25,25) R ^
MSE
0.17236
6.9677e-10
0.10592
4.7716e-07
0.17598
9.5917e-11
0.13517
1.5864e-07
SH1
(75,25,25) R ^
MSE
0.19945
5.9768e-08
0.11034
4.1811e-07
0.17960
2.119e-09
0.13528
1.5773e-07
SH1
(100,25,25) R ^
MSE
0.19819
5.3782e-08
0.10900
4.3554e-07
0.18023
2.733e-09
0.15131
5.6115e-08
SH1
(25,50,25) R ^
MSE
0.17935
1.8924e-09
0.12105
2.9108e-07
0.17843
1.1757e-09
0.15127
5.6325e-08
SH1
(50,75,50) R ^
MSE
0.18495
9.9087e-09
0.11786
3.2649e-07
0.17604
1.0817e-10
0.15383
4.4799e-08
SH1
(75,100,75) R ^
MSE
0.17638
1.9162e-10
0.13235
1.8189e-07
0.1747
8.7569e-12
0.15697
3.2494e-08
SH1
(25,100,25) R ^
MSE
0.19069
2.4633e-08
0.11492
3.6099e-07
0.17653
2.3519e-10
0.1392
1.2816e-07
SH1
(25,25,50) R ^
MSE
0.23891
4.0841e-07
0.12393
2.6079e-07
0.1901
2.2813e-08
0.15131
5.611e-08
SH1
(50,50,75) R ^
MSE
0.16162
1.7909e-08
0.1557
3.7239e-08
0.17367
1.7809e-10
0.16523
9.5422e-09
SH1
(75,75,100) R ^
MSE
0.15119
5.6692e-08
0.13748
1.408e-07
0.17433
4.4538e-11
0.16805
4.8245e-09
SH1
R= 0.16667,   ζ 1 =1.5, ζ 2 =1.5, ζ 3 =1.5and q=10000
Method
n,m,w
Mle Mom SH1 SH2 Best
(25,25,25) R ^
MSE
0.15676
9.82e-09
0.18541
3.5124e-08
0.16392
7.5502e-10
0.17445
6.0599e-09
SH1
(50,50,50) R ^
MSE
0.1404
6.8985e-08
0.21087
1.9543e-07
0.16415
6.3217e-10
0.18916
5.0602e-08
SH1
(75,75,75) R ^
MSE
0.1454
4.3016e-08
0.17792
1.2655e-08
0.16618
2.3994e-11
0.17245
3.3414e-09
SH1
(100,100,100) R ^
MSE
0.17169
2.522e-09
0.16521
2.1217e-10
0.1667
1.3565e-13
0.16773
1.1396e-10
SH1
(50,25,25) R ^
MSE
0.12897
1.4211e-07
0.29976
1.7713e-06
0.15687
9.5912e-09
0.23538
4.7217e-07
SH1
(75,25,25) R ^
MSE
0.1175
2.4171e-07
0.20579
1.5307e-07
0.14959
2.9154e-08
0.18288
2.6288e-08
SH1
(100,25,25) R ^
MSE
0.16667
3.7084e-08
0.14741
4.9826e-07
0.23725
4.7221e-09
0.21319
2.1646e-07
SH1
(25,50,25) R ^
MSE
0.18239
2.4725e-08
0.12568
1.6795e-07
0.17596
8.6408e-09
0.14649
4.072e-08
SH1
(50,75,50) R ^
MSE
0.16269
1.5778e-09
0.15295
1.8804e-08
0.1671
1.907e-11
0.15776
7.9395e-09
SH1
(75,100,75) R ^
MSE
0.13608
9.3539e-08
0.20373
1.374e-07
0.1662
2.1631e-11
0.18712
4.1833e-08
SH1
(25,100,25) R ^
MSE
0.16371
8.7604e-10
0.2123
2.0823e-07
0.1644
5.1378e-10
0.19272
6.7874e-08
SH1
(25,25,50) R ^
MSE
0.13021
1.3291e-07
0.22961
3.9616e-07
0.15341
1.7581e-08
0.2007
1.1581e-07
SH1
(50,50,75) R ^
MSE
0.1759
8.5328e-09
0.13769
8.3974e-08
0.16607
3.5792e-11
0.14916
3.0655e-08
SH1
(75,75,100) R ^
MSE
0.15083
2.5079e-08
0.18493
3.3366e-08
0.1664
7.1447e-12
0.17714
1.0978e-08
SH1
R= 0.1,   ζ 1 =1.5, ζ 2 =3.5, ζ 3 =2.5and q=10000
Method
n,m,w
Mle Mom SH1 SH2 Best
(25,25,25) R ^
MSE
0.1439
1.9272e-07
0.18133
6.6146e-07
0.11091
1.1897e-08
0.17023
4.9322e-07
SH1
(50,50,50) R ^
MSE
0.13639
1.3241e-07
0.14832
2.3344e-07
0.10279
7.8026e-10
0.1621
3.8563e-07
SH1
(75,75,75) R ^
MSE
0.10428
1.8297e-09
0.20421
1.0859e-06
0.099843
2.4697e-12
0.18633
7.4528e-07
SH1
(100,100,100) R ^
MSE
0.10724
5.2398e-09
0.18486
7.2017e-07
0.099781
4.8049e-12
0.18265
6.8303e-07
SH1
(50,25,25) R ^
MSE
0.11243
1.5447e-08
0.18481
7.1931e-07
0.10198
3.9077e-10
0.17709
5.9426e-07
SH1
(75,25,25) R ^
MSE
0.11521
2.3139e-08
0.17945
6.3123e-07
0.10323
1.0435e-09
0.17097
5.0368e-07
SH1
(100,25,25) R ^
MSE
0.10514
2.6379e-09
0.16091
3.7101e-07
0.10201
4.033e-10
0.16608
4.3671e-07
SH1
(25,50,25) R ^
MSE
0.10903
8.1614e-09
0.18471
7.1758e-07
0.10377
1.42e-09
0.19206
8.4754e-07
SH1
(50,75,50) R ^
MSE
0.1112
1.2543e-08
0.21963
1.4312e-06
0.10171
2.9161e-10
0.19207
8.4765e-07
SH1
(75,100,75) R ^
MSE
0.10183
3.3388e-10
0.21353
1.289e-06
0.099982
3.1663e-14
0.18867
7.8616e-07
SH1
(25,100,25) R ^
MSE
0.12711
7.3519e-08
0.19176
8.4195e-07
0.11079
1.1645e-08
0.17787
6.063e-07
SH1
(25,25,50) R ^
MSE
0.12573
6.6195e-08
0.19876
9.7531e-07
0.10639
4.0889e-09
0.18526
7.2684e-07
SH1
(50,50,75) R ^
MSE
0.11281
1.6403e-08
0.1895
8.0111e-07
0.10081
6.6176e-11
0.18458
7.1541e-07
SH1
(75,75,100) R ^
MSE
0.13862
1.4917e-07
0.15759
3.317e-07
0.10038
1.4333e-11
0.16339
4.0187e-07
SH1
R= 0.11667,   ζ 1 =3.5, ζ 2 =2.5, ζ 3 =1.5and q=10000
Method
n,m,w
Mle Mom SH1 SH2 Best
(25,25,25) R ^
MSE
0.11541
1.5718e-10
0.25489
1.9104e-06
0.11629
1.4074e-11
0.21104
8.9072e-07
SH1
(50,50,50) R ^
MSE
0.13119
2.1082e-08
0.22845
1.2496e-06
0.11754
7.7031e-11
0.20084
7.0857e-07
SH1
(75,75,75) R ^
MSE
0.10116
2.4034e-08
0.25097
1.8038e-06
0.116
4.4505e-11
0.21287
9.256e-07
SH1
(100,100,100) R ^
MSE
0.11467
3.9876e-10
0.21453
9.5781e-07
0.1164
7.0629e-12
0.19455
6.0656e-07
SH1
(50,25,25) R ^
MSE
0.12773
1.2229e-08
0.22401
1.1522e-06
0.11957
8.4428e-10
0.19822
6.6501e-07
SH1
(75,25,25) R ^
MSE
0.14116
5.9999e-08
0.18953
5.3089e-07
0.12366
4.8937e-09
0.18391
4.522e-07
SH1
(100,25,25) R ^
MSE
0.13117
2.1028e-08
0.20523
7.8439e-07
0.12032
1.334e-09
0.19259
5.7639e-07
SH1
(25,50,25) R ^
MSE
0.11238
1.8361e-09
0.18634
4.855e-07
0.11584
6.8685e-11
0.17234
3.0991e-07
SH1
(50,75,50) R ^
MSE
0.12615
8.9875e-09
0.2232
1.1349e-06
0.11812
2.1166e-10
0.19613
6.3138e-07
SH1
(75,100,75) R ^
MSE
0.11136
2.8167e-09
0.23389
1.3742e-06
0.11633
1.1273e-11
0.20431
7.682e-07
SH1
(25,100,25) R ^
MSE
0.13305
2.6836e-08
0.20444
7.7047e-07
0.1194
7.4595e-10
0.18279
4.3717e-07
SH1
(25,25,50) R ^
MSE
0.13803
4.5657e-08
0.15661
1.5957e-07
0.11943
7.6543e-10
0.16303
2.1493e-07
SH1
(50,50,75) R ^
MSE
0.12893
1.5035e-08
0.22632
1.2024e-06
0.117
1.1364e-11
0.19254
5.7566e-07
SH1
(75,75,100) R ^
MSE
0.13302
2.6751e-08
0.19233
5.7249e-07
0.11656
1.0731e-12
0.18769
5.0448e-07
SH1
R= 0.23333,   ζ 1 =3.5, ζ 2 =1.5, ζ 3 =2.5and q=10000
Method
n,m,w
Mle Mom SH1 SH2 Best
(25,25,25) R ^
MSE
0.21429
3.6267e-08
0.13464 9.7408e-07 0.22908
1.8092e-09
0.15108
6.7651e-07
SH1
(50,50,50) R ^
MSE
0.19785
1.2591e-07
0.11743
1.3433e-06
0.2306
7.4572e-10
0.13693
9.2927e-07
SH1
(75,75,75) R ^
MSE
0.21823
2.2806e-08
0.10601
1.6212e-06
0.23325
6.592e-13
0.13887
8.9235e-07
SH1
(100,100,100) R ^
MSE
0.21461
3.5065e-08
0.12208
1.2378e-06
0.23348
2.1933e-12
0.15091
6.7935e-07
SH1
(50,25,25) R ^
MSE
0.2452
1.4077e-08
0.11016
1.5172e-06
0.23601
7.1686e-10
0.14458
7.8772e-07
SH1
(75,25,25) R ^
MSE
0.17944
2.9049e-07
0.14782
7.3125e-07
0.2162
2.9371e-08
0.1579
5.6829e-07
SH1
(100,25,25) R ^
MSE
0.26793
1.1967e-07
0.10931
1.5382e-06
0.24479
1.3124e-08
0.14556
7.705e-07
SH1
(25,50,25) R ^
MSE
0.18239
2.595e-07
0.10106
1.7496e-06
0.22039
1.6755e-08
0.13968
8.7702e-07
SH1
(50,75,50) R ^
MSE
0.22786
2.9949e-09
0.10771
1.5781e-06
0.23371
1.3946e-11
0.13791
9.106e-07
SH1
(75,100,75)
R ^
MSE
0.24163
6.878e-09
0.10018
1.7731e-06
0.23389
3.0849e-11
0.13539
9.5927e-07
SH1
(25,100,25)
R ^
MSE
0.188
2.0554e-07
0.13646
9.3851e-07
0.22451
7.7868e-09
0.15624
5.943e-07
SH1
(25,25,50) R ^
MSE
0.29333
3.6e-07
0.10743
1.5852e-06
0.24206
7.6197e-09
0.14321
8.1225e-07
SH1
(50,50,75) R ^
MSE
0.18978
1.8966e-07
0.14026
8.6626e-07
0.23203
1.6882e-10
0.1577
5.7203e-07
SH1
(75,75,100) R ^
MSE
0.22703
3.9738e-09
0.10696
1.597e-06
0.23361
7.6452e-12
0.13775
9.1362e-07
SH1

Co-author Information

References

  1. Hassan A. S. and Basheikh H. M., (2012), "Estimation of Reliability in Multi-Component Stress-Strength Model Following Exponential Pareto Distribution", The Egyptian Statistical Journal, Institute of Statistical Studies Research, Cairo University, Vol. 56, No.2, pp. 82-95.
  2. S. Kotz, Y. Lumelskii, and M. Pensky, The stress- strength model and its generalizations, “Theory and Applications” (World Scientific Publishing Co., 2003). [CrossRef]
  3. Isaam, K, Taha, A. and Abbas,N." Different Estimation Methods of Reliability in Stress -Strength Model under Chen Distribution" AIP Conference Proceedings 2591, 050023, 1-10, 2023.
  4. S. Chandra and D. B. Owen, On estimating the reliability of a component subject to several different stresses (strengths). Naval Research Logistics Quarterly, 22, 1975, 31-39. [CrossRef]
  5. N. Singh, On the estimation of of P=R(X1 < Y < X2). Communication in Statistics Theory & Methods, 9, 1980, 1551-1561. [CrossRef]
  6. K. Dutta, and G. L. Sriwastav, An n-standby system with P(X < Y < Z). IAPQR Transaction, 12, 1986, 95-97.
  7. V. V. Ivshin, On the estimation of the probabilities of a double linear inequality in the case of uniform and two-parameter exponential distributions, Journal of Mathematical Science, 88, 1998, 819-8. [CrossRef]
  8. Z. Wang, W. Xiping, and P. Guangming, "Nonparametric statistical inference for P(X <Y <Z)". The Indian Journal of Statistics, 75-A (1), 2013, 118-138. [CrossRef]
  9. M. Ahsanullah, Record values, "The Exponential Distribution: Theory, Methods and Applications", N. Balakrishnan and A. P. Basu, eds., Gordon and Breach Publishers, New York, New Jersey, 1995b.
  10. C.W. Topp and F.C. Leone."A family of J-shaped frequency functions", J.Amer. Statist.Assoc. 50 (1955), pp. 209–219. [CrossRef]
  11. S. Nadarajah and S. Kotz."Moments of some J-shaped distributions", J. Appl. Statist. 30 (2003), pp. 311–317. [CrossRef]
  12. S. Kotz and J.R. Van Dorp, "Beyond Beta: Other Continuous Families of Distributions with Bounded Support and Applications", World Scientific, Singapore, 2004.
  13. M.E. Ghitany, S. Kotz, and M. Xie." On some reliability measures and their stochastic orderings for the Topp–Leone distribution", J. Appl. Statist. 32 (2005), pp. 715–722. [CrossRef]
  14. TAHA .A "On Reliability Estimation for the RayleighDistribution Based on Monte Carlo Simulation" International Journal of Science and Research (IJSR) ISSN (Online): 2319-7064, 2015.
  15. Thompson, J.R."Some Shrinkage Techniques for Estimating the Mean". J. Amer. Statist. Assoc., 63, 113-122 ,1968. [CrossRef]
  16. Isaam, K, Taha, A. and Abbas,N." Estimation of (S-S) Reliability for Inverted Exponential Distribution" AIP Conference Proceedings 2591, 050006,1-11, 2023. [CrossRef]
  17. A.N.Salman "Pre-test single and double stage shrunken estimators for the mean of normal distribution with known variance". Baghdad Journal for Science, 7(4), (1432–1442),2010.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
Prerpints.org logo

Preprints.org is a free preprint server supported by MDPI in Basel, Switzerland.

Subscribe

© 2024 MDPI (Basel, Switzerland) unless otherwise stated