ALL Metrics
-
Views
-
Downloads
Get PDF
Get XML
Cite
Export
Track
Research Article

A new modified Lehmann type – II G class of distributions: exponential distribution with theory, simulation, and applications to engineering sector

[version 1; peer review: 1 approved with reservations, 1 not approved]
PUBLISHED 17 Jun 2021
Author details Author details
OPEN PEER REVIEW
REVIEWER STATUS

Abstract

Background: Modeling against non-normal data a challenge for theoretical and applied scientists to choose a lifetime model and expect to perform optimally against experimental, reliability engineering, hydrology, ecology, and agriculture sciences, phenomena.
Method: We have introduced a new G class that generates relatively more flexible models to its baseline and we refer to it as the new modified Lehmann Type – II (ML–II) G class of distributions. A list of new members of ML–II-G class is developed and as a sub-model the exponential distribution, known as the ML-II-Exp distribution is considered for further discussion. Several mathematical and reliability characters along with explicit expressions for moments, quantile function, and order statistics are derived and discussed in detail. Furthermore, plots of density and hazard rate functions are sketched out over the certain choices of the parametric values. For the estimation of the model parameters, we utilized the method of maximum likelihood estimation.
Results: The applicability of the ML–II–G class is evaluated via ML–II–Exp distribution. ML–II–Exp distribution is modeled to four suitable lifetime datasets and the results are compared with the well-known competing models. Some well recognized goodness–of–fit including -Log-likelihood (-LL), Anderson-Darling (A*), Cramer-Von Mises (W*), and Kolmogorov-Smirnov (K-S) test statistics are considered for the selection of a better fit model.
Conclusion: The minimum value of the goodness–of–fit is the criteria of a better fit model that the ML–II–Exp distribution perfectly satisfies. Hence, we affirm that the ML–II–Exp distribution is a better fit model than its competitors.

Keywords

Lehmann Type – II Distribution; Exponential Distribution; Hazard Rate Function; Moments; Rényi Entropy; Maximum Likelihood Estimation, Simulation.

Introduction

Over the past two decades, the growing attention of researchers towards the development of new G families has explored the remarkable characteristics of baseline models. New models open new horizons for theoretical and applied researchers to address real-world problems, to proficiently and adequately fit them to asymmetric and complex random phenomena. Accordingly, several classes of distributions have been developed and discussed in the literature. For more details, we encourage the reader to see the credible work of some notable scientists including Marshall and Olkin1 Gx/Gx+αG¯x, who developed a new technique to generate models; Eugene et al.,2 who proposed beta generated-G class Gbetax; the quadratic rank transmutation map technique introduced by Shaw and Buckley3 [1+λGxλG2x]; the Kumaraswamy generalized-G class proposed by Cordeiro and Castro4 [11Gαxβ]; the gamma-G class proposed by Ristic and Balakrishnan5 logGx; the exponentiated generalized (EG) class proposed by Cordeiro et al.6 [11Gxαβ]; the T–X family proposed by Alzaatreh et al.7 1RWGx; the Weibull–G family proposed by Bourguignon et al.8 [1exp(αGx/G¯xβ)]; the beta Marshall–Olkin–G class proposed by Alizadeh et al.9 IMarshallOlkinab; the logistic-X family proposed by Tahir et al.10 [1+logG¯xα1]; Kumar et al.11 proposed DUS transformation eGx1/e1; Mahdavi and Kundu12 proposed Alpha transformation αGx1/α1; Elbatal et al.13 proposed a new Alpha power transformed family of distributions GxαGx/α; and recently, Ijaz et al.14 proposed a Gull alpha power Weibull family αGx/αGx; among others.

This paper is organised into the following sections. A new modified Lehmann Type–II (ML–II) G class of distributions accompanied by a table of some special models is proposed and developed in A new modified Lehmann Type – II–G Class of distributions. A special model of ML–II–G class, known as a modified Lehmann Type–II exponential (ML–II–Exp) distribution, along with its mathematical and reliability measures are derived and discussed in Mathematical properties. The method of maximum likelihood estimation is used to estimate the unknown model parameters and develop some simulation results to assess the performance of maximum likelihood estimations (MLEs) in Inference. Applications are discussed in four real data applications and finally, the conclusion is reported in Conclusions.

A new modified Lehmann Type – II–G Class of distributions

Lehmann15 proposed the Lehmann Type–I (L - I) distribution, which was the simple exponentiated version of any arbitrary baseline model. Accordingly, the first credit goes to Gupta et al.16 because they applied L – I on exponential distribution. The associated cumulative distribution function (CDF) is given by

Fxαξ=Gαxξ

Cordeiro et al.4 deserves to be acknowledged as they use dual transformation that yielded the Lehmann Type – II (L–II) G class of distributions. The associated CDF is given by

Fxαξ=11Gxξα

The closed-form feature of the L–II distribution assists one to derive and study its numerous properties and in literature, both the approaches (L–I and L–II) have been extensively utilized to study the unexplored characteristics of the classical and modified models.

We develop a new G class, known as a modified Lehmann Type–II (ML–II) G class of distributions. The corresponding CDF is given by

(1)
FMLIIGxαβξ=11Gxξ1αGxξβ,xϵ

where Gxξϵ01 is a CDF of any arbitrary baseline model based on the parametric vector ξ, dependent on (r x 1) with <α<1, and α,β>0 are the scale and shape parameters, respectively. Let gxξ=dGxξ/dx be the density function of any baseline model. The associated probability density function (PDF), hazard rate function (HRF), and quantile function of ML–II–G class of distributions are given by, respectively

(2)
fMLIIGxαβξ=β1αgxξ1Gxξβ11αGxξ1+β
(3)
hMLIIGxαβξ=β1αgxξ1Gxξβ11αGxξ1+2β1αGxξβ1Gxξβ
(4)
QMLIIGpαβξ=G11p1β1α1p1β1,p01

Now and onward, the modified Lehmann Type–II–G random variable X corresponding to fMLIIGxαβξ will be denoted by X~ML– II–Gxαβξ.

This study aimed to propose a new G class of distributions that generates more flexible alternative continuous models relative to its parent distribution. From a computational point of view, new models are very simple to interpret. New models offer greater distributional flexibility and can provide a better fit over the complex random phenomena that exclusively arise in engineering sciences. However, to the best of our knowledge, no study has been conducted previously that discusses our new G class, deliberates the unexplained complex random phenomena so well and advances the fit to a diverse range of sophisticated lifetime data.

Linear combination provides a much more informal approach to discuss the CDF and PDF than the conventional integral computation when determining the mathematical properties. For this, binomial expansion is given as follows:

1zβ=i=01iβizi,z<1

Infinite linear combinations of CDF (1) and PDF (2) of ML–II–G class are given by, respectively

Fxαβξ=1i,j=01i+jαjβiβjGi+jxαβξ
(5)
Fxαβξ=1i,j=0ωijGi+jxαβξ
fxαβξ=β1αi,j=01i+jβ1iβ1jGi+jxαβξgxαβξ
(6)
fxαβξ=β1αi,j=0ωijGi+jxαβξgxαβξ

The r-th ordinary moment (say μr/) of X is given by

(7)
μr,I/=β1αi,j=0ωijIr,i+jxξ

where coefficient ωij=1i+jαjβiβj, ωij=i,j=01i+jβ1iβ1j, and Ir,i+jxξ=xrGi+jxξgxξdx.

Further, by (1), Lehmann Type–II–G class and baseline Gx model can be traced back at α=0, and α=0,β=1, respectively. The expansions of CDF (5) and PDF (6) provide us with the Exp-G class, which is quite useful for the generalization of models.

Order statistics

In reliability analysis and life testing of a component in quality control, order statistics (OS) and its moments are considered as a noteworthy measure. Let X1,X2,,Xn be a random sample of size n that follows the Lehmann Type–II–G class and X1:n<X2:n<<Xn:n be the corresponding OS. The random variables Xi,X1, and Xn be the i-th, minimum, and maximum OS of X.

The PDF of Xi is given by

fi:nxξ=1Bini+1!Fxξi11Fxξnifxξ

where =1,2,3,,n,Fxξ, and fxξ are the associated CDF (5) with corresponding PDF (6) of the Lehmann Type–II–G family. Using the fact that

1Fxξni=m=0ni1mnimFxξm

By placing the last expression in fi:nxξ, we get the most refined form of OS PDF and one may determine it by integrating (5–6) and the expression may be given as follows

fi:nxξ=fxξBini+1!m=0ni1mnimFxξi+m1

A modified Lehmann Type – II Exponential (Ml–II–Exp) distribution (sub-model)

In this section, we introduce a sub-model of ML–II–G class of distributions, known as a ML–II–Exp distribution. For this, we have the CDF and PDF of the exponential distribution GExpxθ=1eθx, and gExpxθ=θeθx,forθ>0,0<x<, respectively. The associated CDF and corresponding PDF of the ML–II–Exp distribution are obtained by following (1–2) and its analytical expressions are given by respectively:

(8)
FMLIIExpxαβθ=1eθx1α+αeθxβ,<α<1
(9)
fMLIIExpxαβθ=1αβθeθβx1α+αeθxβ+1

where α is a scale and β,θ>0 are the shape parameters, respectively. By following (1–2), linear representation of CDF and PDF are given as follows, respectively

(10)
Fxαβθ=1i,j=01jθjβ+ijβiαj1αβij!xj
Fxαβθ=1i,j=0ζijxj
(11)
fxαβθ=βθi=0αi1αββ1iei+βθx
(12)
fxαβθ=βi,j=01j1αβiαiθj+1β1iβ+ijj!xj
fxαβθ=βi,j=0ζijxj.

where ζij=1jθjβ+ijβiαj1αβij!, ζij=1j1αβiαiθj+1β1ii+βjj!

Expression in (11) is expected to be quite supportive in the forthcoming computations of various mathematical properties of the ML–II–Exp distribution.

Mathematical properties

Reliability characteristics

One of the imperative roles of probability distribution in reliability engineering is to analyze and predict the life of a component. One may define the reliability function as the probability that a component survives until the time x and analytically it can be written as Rx=1Fx.

The reliability function of X is given by

(13)
RML-II-Expxαβθ=eθx1α+αeθxβ

In reliability theory, the significant contribution of a function that measures the failure rate of a component in a particular time t is sometimes referred to as the HRF, failure rate function, or the force of mortality, and mathematically it can be written as hx=fx/Rx.

The HRF of X is given by

(14)
hMLIIExpxαβθ=1αβθ1α+αeθx

Numerous notable reliability measures for the ML–II–Exp distribution can be discussed and derived, such as reverse HRF by hrxαβθ=fxαβθ/Rxαβθ, Mills ratio by Mxαβθ=Rxαβθ/fxαβθ, and Odd function by

Oxαβθ=Fxαβθ/Rxαβθ.

Shapes

Different plots of PDF and HRFs of the ML–II–Exp distribution are sketched over the selected and fixed combinations of the model parameters, respectively. Figure 1 (a, b, c) presents the reversed-J, constant, unimodal, and right-skewed shapes of the PDF and Figure 2 (a, b, c) illustrates the decreasing and increasing HRF. However, an increasing HRF with some interesting facts are identified when suddenly spikes arise at the tail end of HRF is unexpectedly detected. Such kinds of trends are often observed in non-stationary time series lifetime phenomena.

Table 1. Some special models and corresponding G(x, ξ) and S(x, ξ).

ModelG(x)S(x)ξ
Rayleigh
(x>0)
1eηx2eηx21α1eηx2βα,β,η
Gompertz
(x>0)
1eηeγx1eηeγx11+α1eηeγx1βα,β,η,γ
Pareto
(x>m)
1mxηmxη1α1mxηβα,β,η
Fréchet
(x>0)
eηxγ1eηxγ1αeηxγβα,β,η,γ
Burr X
(x>0)
1eηx2γ11eηx2γ1α1eηx2γβα,β,η,γ
Weibull
(x>0)
1eηxγeηxγ1α1eηxγβα,β,η,γ
Lomax
(x>0)
11+xηγ1+xηγ1α11+xηγβα,β,η,γ
6d110898-0422-498a-a5a4-5ef5ca29e74e_figure1.gif

Figure 1. Density function plots of ML–II–Exponential distribution.

6d110898-0422-498a-a5a4-5ef5ca29e74e_figure2.gif

Figure 2. Hazard rate function plots of ML–II–Exponential distribution.

Limiting behavior

Here we study the limiting behavior of CDF, PDF, reliability, and HRFs of the ML–II–Exp distribution present in (8), (9), (13), and (14) at x0 and x.

Proposition-1: Limiting behavior of CDF, PDF, reliability, and HRFs of the ML–II–Exp distribution at x0 is followed by

F0αβθ0f0αβθ1αβθR0αβθ1h0αβθ1αβθ

Proposition-2: Limiting behavior of CDF, PDF, reliability, and HRFs of the ML–II–Exp distribution at x is followed by

Fαβθ1fαβθ0Rαβθ0hαβθ0

Limiting behaviors developed in the above expressions may illustrate the effect of parameters on the tail of the ML–II–Exp distribution.

Moments and associated measures

Moments have a remarkable role in the discussion of the distribution theory, to study the significant characteristics of a probability distribution such as mean; variance; skewness, and kurtosis.

Theorem 1: If X ML–II–Exp (x;α,β,θ), for x;β,θ>0, with <α<1, then the r-th ordinary moment (say μr/) of X is given by

μr/=βθri=0ηiβ+ir+1Γr+1

where ηi=αi1αβiβ1i

Proof: μr/ can be written by following (11), as

(15)
μr/=βθi=0αi1αβiβ1i0xreθxβ+idx

Let’s suppose θxβ+i=y;x=yθβ+idx=dyθβ+i

limits: as x0x;y0y.

By placing the above information in (15), we get

μr/=βθi=0αi1αβiβ1i0yθβ+ireyθβ+idy

by making simple computation on the last expression leads us to the r-th ordinary moment, in terms of the gamma function and it is given by

(16)
μr/=βθri=0ηiβ+ir+1Γr+1,r=1,2,n

where Γ. is a gamma function,Γx=0tx1etdt,ηi=αi1αβiβ1i, and <α<1.

The derived expression in (16) may serve a supportive and useful role in the development of several statistical measures. For instance: to deduce the mean of X, placer = 1 in (16). For the higher-order ordinary moments of X approximating to 2nd, 3rd, and 4th, and higher, these moments can be formulated by setting r = 2, 3, and 4 in (16), respectively. Additionally, to discuss the variability in X, the Fisher index (F. I = (VarX/EX)) may play a significant role in this scenario. To derive the negative moments of X, simply substitute r by –w in (16). One may perhaps further determine the well-established statistics such as skewness (τ1=μ32/μ23), and kurtosis (τ2=μ4/μ22), of X by integrating (16). A well-established relationship between the central moments μs and cumulants (Ks) of X may easily be defined by ordinary moments μs=k=0ssk1kμ1/sμsk/. Hence, the first four cumulants can be calculated by K1=μ1/,K2=μ2/μ1/2 , K3=μ3/3μ2/μ1/+2μ1/3, and K4=μ4/4μ3/μ1/3μ2/2+12μ2/μ1/26μ1/4, etc., respectively.

In Table 2, some numerical results of the first eight ordinary moments, ε2 = variance, τ1= skewness, and τ2= kurtosis for some chosen parameters are presented in S-I (α=0.9,θ=0.1,β=1.1), S-II (α=0.9,θ=0.5,β=1.5), S-III (α=0.8,θ=0.09,β=3.1), and S-IV (α=0.5,θ=0.009,β=5.1).

Table 2. Some numerical results of moments, variance, skewness, and kurtosis.

μ/rS-IS-IIS-IIIS-IV
μ/11.06691.09530.43280.3951
μ/22.21842.06240.36820.3119
μ/36.82585.33060.46380.3688
μ/427.802617.38490.77140.5811
μ/5141.017968.42751.59201.1437
μ/6856.5819316.01553.92042.6994
μ/76063.83001678.429011.21597.4292
μ/849031.230010095.810036.552623.3579
ε21.94271.72610.25780.2129
τ12.95961.82251.99872.0879
τ25.03913.35524.50274.7385

Residual and reversed residual life functions

The residual life function/conditional survivor function of random variable Rt=X+t/X>t X is the probability that a component whose life says x, survives in an additional interval at t0. Analytically it can be written as

SRtx=Sx+tSt

The residual life function of X is given by

(17)
SRtMLIIExpxαβθ=eθβx+t1α+αeθtβeθβt1α+αeθx+tβ,x>0

with associated CDF

(18)
FRtMLIIExpxαβθ=eθβt1α+αeθx+tβeθβx+t1α+αeθtβeθβt1α+αeθx+tβ

PDF and HRF corresponding to (18) are given as follows, respectively

(19)
fRtMLIIExpxαβθ=βθ1α1α+αeβeβθx+t+θx+t+βtθ1α+αeθx+tβα+1αeθx+t
(20)
hRtMLIIExpxαβθ=βθ1αeβθx+t1α+αeθx+t

Mean residual life function is given by

ESRx=1Stμ1/0txfxdxt,t0

Moreover, the reverse residual life can be defined as: R¯t=tX/Xt.

SR¯tx=FtxFt,t0

The reverse residual life function of X is given by

(21)
SR¯tMLIIExpxαβθ=1α+αeθtxβeβθtx1α+αeθtβ1α+αeθtβeβθt1α+αeθtxβ

with associated CDF

(22)
FR¯tMLIIExpxαβθ=1α+αeθtβeβθt1α+αeθtxβ1α+αeθtxβeβθtx1α+αeθtβ1α+αeθtβeβθt1α+αeθtxβ

PDF and HRF corresponding to (22) are given as follows, respectively

(23)
fR¯tMLIIExpxαβθ=βθ1α1α+αeβeβθtx+θtx+βtθ11α+αeβeβtθ1α+αeθtxβα+1αeθtx
(24)
hR¯tMLIIExpxαβθ=βθ1αeβθtx1α+αeθtx

Mean reversed residual life function/mean waiting time is given by

ESR¯x=t1Ft0txfxdx,t0

One may derive the strong mean inactivity time of X by following

Mt=t21ft0tx2fxdxfort0

where Φrt=0txrfxdx, is the r–th incomplete moment, μ1t/=0txfxdx=βθi=0ηiβ+i2Γt2, and μ2t/=0tx2fxdx=βθ2i=0ηiβ+i3Γt3. By following (16), we may derive directly the μ1t/ and μ2t/ by holding t=β+i as the upper bound. Furthermore, μ1t/ and μ2t/ are termed as the first and second lower incomplete moments of X, respectively, with ηi=αi1αβiβ1i, and <α<1.

Entropy

When a system is quantified by disorderedness, randomness, diversity, or uncertainty, in general, it is known as entropy.

Rényi17 entropy of X is given by

(25)
HζX=11ζlog0fζxdx,ζ>0andζ1

By following (9), we simplify fx in terms of fζx, we get

Hζ-ML-II-ExpX=1αβθζ1αζβ+1eβθζx1+α1αeθxζβ+1

by placing the above expression in (25), we get

Hζ-ML-II-ExpX=11ζlog1αβθζ1αζβ+10eβθζx1+α1αeθxζβ+1dx
Hζ-ML-II-ExpX=11ζlogβθζ1αζβj=0ζβ+1jα1αj0eθζβ+1xdx

hence, solving simple mathematics on the previous equation leads us to the most simplified form of the Rényi entropy of X and it is given by

Hζ-ML-II-ExpX=11ζlogτζj=0vj

where vj=α1αjζβ+1j,τζ=1θζβ+1,<α<1.

Quantile function, and mode

The qth quantile function of ML–II–Exp distribution is obtained by inverting the CDF. Quantile function is defined as q=Fxq=PXxq,q01.

The quantile function of X is given by

(26)
xq-ML-II-Exp=1θlog1α1q1β1α1q1β

To obtain the 1st quartile, median and 3rd quartile of X, place q = 0.25, 0.5, and 0.75 respectively in (26). Henceforth, to generate random numbers, one may assume that the CDF in (8) follows the uniform distribution u = U (0, 1).

The modal value of X is calculated by following the constraint f/xαβθ=dfxαβθdx=0. For convenience, fxαβθ can be rewritten as

fML-II-Expxαβθ=1αβθeβθx1α+αeθxβ+1

The simplified form of f/xαβθ is given by

f/ML-II-Expxαβθ=αeθx1αβ1α+αeθxβ+1

Hence, solving simple algebra on the previous equation may provide us with the most suitable form of the mode of X in support of f/xαβθ=0 and it is given by

x̂ML-II-Exp=1θlogαβ1α,<α<1.

Stress – strength reliability

Let X1 and X2 be defined to discuss the strength and stress of a component, respectively, followed by the same uni-variate family of distributions, which will work in order if X2<X1. To discuss the reliability (say R) of X, it is given by R=PX2<X1.

Theorem 2: Let X1 ML–II–Exp (x;α,β1,θ) and X2 ML–II–Exp (x;α,β2,θ) be independent random variables following the ML–II–Exp distribution; then the reliability is given by

R=β1β1+β2

Proof: Reliability (R) is defined as

R=PX2<X1=0f1xF2xdx.

R of X can be written by following (9), as

(27)
RML-II-Exp=01αβ1θeθβ1x1α+αeθxβ1+11eθx1α+αeθxβ2dx,

Let’s suppose t=eθx1α+αeθxβ1dt=1αβ1θeθβ1x1α+αeθxβ1+1dx;andtβ2β1=eθx1α+αeθxβ2

limits: as x0t1;xt0.

By placing the above information in (27), we have

RML-II-Exp=011tβ2β1dt,

Hence, the simple computation of the above expression provides us with the reduced form of R in terms of β1 and β2, as we presume that the R is a function of β1 with increasing behavior and it is given by

RML-II-Exp=β2β1+β2

Order statistics

In reliability analysis and life testing of a component in quality control, order statistics OS and its moments are considered as a noteworthy measure. Let X1,X2,,Xn be a random sample of size n following the ML–II–Exp distribution and X1:n<X2:n<<Xn:n be the corresponding OS. The random variables Xi,X1, and Xn be the i-th, minimum, and maximum OS of X.

The PDF of Xi is given by

fi:nx=1Bini+1!Fxi11Fxnifx,i=1,2,3,,n

By following (8) and (9), the PDF of Xi takes the form

(28)
fi:n-ML-II-Expxαβθ=1Bi,ni+1!1eθx1α+αeθxβi1eθx1α+αeθxβni1αβθeθβx1α+αeθxβ+1

by utilizing some the techniques of binomial expansion (mentioned in the new modified Lehmann Type – II–G Class of distributions) to simplify (28), we get the reduced form of fi:nxαβθ and it is given as follows

(29)
fi:n-ML-II-Expx=βθBi,ni+1!j=0k=0ηjkeθxv

and we determine the linear representation of (29) and it can be written as

(30)
fi:nML-II-Expx=βBi,ni+1!j=0k=0l=0jklxl

Indeed, (29) has a supportive role in the calculation of r-th moment OS and hereafter, straightforward computation of (29) leads us to the r-th moment OS of X and it is given as follows

(31)
μOS-ML-II-Expr=βBi,ni+1!j=0k=0ηjkθvr+1Γr+1

where ηjk=1ji1jλkαk1αλk,λ=βni+j+1+1,v=βθni+j+1

jkl=1j+ki1jλkvl/l!

The CDF of Xi is given by

Fi:nx=r=innrFxr1Fxnr,i=1,2,3,,n
(32)
Fi:n-ML-II-Expx=r=innr1eθx1α+αeθxβreθx1α+αeθxβnr

Furthermore, the minimum and maximum OS of X follows directly from (28) with i = 1 and i = n, respectively.

Inference

In this section, we estimate the parameters of the ML–II–Exp distribution by following the method of MLE, as this method provides the maximum information about the unknown model parameter. Let X1,X2,X3,,Xn be a random sample of size n from the ML–II–Exp distribution, then the likelihood function LML-II-Expψ=αβθ=i=1nfMLIIExpxiψ of X is given by

LML-II-Expψ=i=1n1αβθeβθxi1α+αeθxiβ+1

The log-likelihood function, lψ is given by

(33)
lML-II-expψ=nlog1α+logβ+logθβθi=1nxiβ+1i=1nlog1α+αeθxi

Partial derivatives of (33) w.r.t. α,θ, and β yield, respectively

lML-II-Expψ∂α=n1αβ+1i=1n1+eθxi1α+αeθxi=0
lML-II-Expψβ=nβθi=1nxii=nlog1α+αeθxi=0
lML-II-Expψ∂θ=nθβi=1nxiβ+1i=1nαθeθxi1α+αeθxi=0

The maximum likelihood estimates ( ψ̂i=α̂,β̂,θ̂) of the ML–II–Exp distribution can be obtained by maximizing (33) or by solving the above non-linear equations simultaneously. These non-linear equations, however, do not provide an analytical solution for the MLEs and the optimum value of α, β and θ. Consequently, iterative techniques such as the Newton-Raphson type algorithm are an appropriate choice in the support of MLEs.

Simulation study

In this sub-section, we discuss the performance of MLEs using the following algorithm.

Step 1: A random sample x1, x2, x3 , ..., xn of sizes n = 25, 50, 100, 200, 300, 400, and 500 are generated from (26).

Step 2: The required results are obtained based on the different combinations of the model parameters placed in S-V (α=0.9,θ=1.1,β=0.5), S-VI (α=0.5,θ=1.2,β=0.6), and S-VII (α=0.9,θ=0.2,β=0.06), S-VIII (α=0.5,θ=0.09,β=0.2), S-IX (α=0.1,θ=0.1,β=0.2) and S-X (α=0.9,θ=0.1,β=0.9).

Step 3: Average MLEs and their corresponding standard errors (SEs) (in parenthesis) are presented in Table 4.

Step 4: Estimated bias, root mean square error (RMSE), variance, and mean values are presented in Table 5.

Step 5: Each sample is replicated N = 500 times.

Step 6: A gradual decrease in SEs, biases, RMSE, variances, means, and MLEs pretty close to the true parameters are observed with increase in the sample sizes.

Step 7: Finally, the estimates present in Tables 4 and 5 help us to specify that the method of maximum likelihood works consistently for the ML–II–Exp distribution.

The in-practice measures for the development of average estimate (AE), SE, bias, and RMSE are given as follows:

AEψ̂=1Ni=1Nψ̂,SEψ̂=1Ni=1Nψ̂iψ¯2,RMSEψ̂=1Ni=1Nψ̂iψ2,
Biasφ̂=1Ni=1Nψ̂iψ,andCIψ̂=ψ̂±Zα2varψ̂

Four real data applications

In this section, we explore four suitable lifetime datasets to model the ML–II–Exp distribution. These datasets are associated with the engineering sector. The first dataset relates to the study of failure times of 84 windshields for a particular model of aircraft (the unit for measurement is 1000 hours) that was first discussed by Ramos et al.18 The second dataset relates to the study of service times of 63 aircraft windshields (the unit for measurement is 1000 hours) that was discussed by Tahir et al.19 The third dataset follows the discussion of the breaking stress of carbon fibers (in Gba) that was initially developed by Nicholas and Padgett20 and finally the fourth one relates to the study of fatigue life of 6061 - T6 aluminum coupons cut parallel to the direction of rolling and oscillated at 18 cycles per second. The data set consists of 101 observations with maximum stress per cycle of 31,000 psi. This dataset was pioneered by Birnbaum and Saunders.21 Subsequently, this data was discussed by Shanker et al.,22 after subtracting 65 from each observation. The datasets are given in the underlying data.

The ML–II–Exp distribution is compared with the well-known models (CDF list is mentioned in Table 6). We follow some recognized selection criteria including -Log-likelihood (-LL), Anderson-Darling (A*), Cramer-Von Mises (W*), and Kolmogorov-Smirnov (K-S) test statistics. Some common results of descriptive statistics such as minimum value, 1st quartile, means, 3rd quartile, 95% confidence interval, and the maximum value, are tabulated in Table 3. The parameter estimates, standard errors (in parenthesis), and the goodness-of-fit are confirmed in Tables 7-10, respectively. The minimum value of the goodness-of-fit is the criteria of the better fit model that the ML–II–Exp distribution perfectly satisfies. Hence, we affirm that the ML–II–Exp distribution is a better fit than its competitors.

Table 3. Descriptive statistics.

DatasetMinimum1st
Quartile
Mean3rd
Quartile
95%
Confidence Interval
Maximum
10.0401.8662.5633.376(2.32,2.80)4.663
20.0461.1222.0852.820(1.77,2.40)5.140
30.3901.8402.6213.220(2.42,2.82)5.560
45.00055.0068.3380.25(63.89,72.77)147.0

Table 4. Average MLEs and standard errors (in parenthesis).

nS-V
Estimates
(Standard Errors)
S-VI
Estimates
(Standard Errors)
S-VII
Estimates
(Standard Errors)
α̂β̂θ̂α̂β̂θ̂α̂β̂θ̂
250.8895
(0.0979)
0.4935
(0.4787)
1.3017
(0.9226)
0.4036
(0.4076)
0.6612
(1.9725)
0.9195
(2.4506)
0.8031
(0.3244)
0.0764
(0.0725)
0.1573
(0.1428)
500.9303
(0.0397)
0.6702
(0.3945)
1.1816
(0.4331)
0.5499
(0.2228)
0.5939
(0.5831)
1.4548
(1.2127)
0.9785
(0.0399)
0.0854
(0.6502)
0.2136
(0.1452)
1000.8997
(0.0408)
0.5616
(0.2859)
1.1073
(0.3819)
0.4599
(0.1847)
0.8215
(0.8319)
1.0985
(0.9456)
0.9696
(0.0535)
0.0597
(0.0481)
0.1929
(0.1445)
2000.9261
(0.0236)
0.4642
(0.1344)
1.2698
(0.2561)
0.3174
(0.1657)
0.4879
(0.5244)
1.2621
(1.2587)
0.8903
(0.0803)
0.0573
(0.0239)
0.1927
(0.0761)
3000.9199
(0.0235)
0.3574
(0.0871)
1.3011
(0.2391)
0.5149
(0.1087)
0.3991
(0.2764)
1.6572
(1.0288)
0.9198
(0.0476)
0.0700
(0.0221)
0.1755
(0.0516)
4000.9098
(0.0224)
0.3528
(0.0772)
1.3579
(0.2252)
0.4738
(0.0886)
0.6453
(0.3588)
1.1241
(0.5378)
0.9415
(0.0391)
0.0552
(0.0183)
0.2169
(0.0678)
5000.9281
(0.0145)
0.4973
(0.0996)
1.1945
(0.1629)
0.5620
(0.0659)
0.7147
(0.3690)
1.0924
(0.4628)
0.9606
(0.0240)
0.0547
(0.0139)
0.2175
(0.0521)

Table 5. Bias, Root Mean Square Error (RMSE), variance, and mean.

nEst.S-VIIIS-IXS-X
αβθαβθαβθ
25Bias0.19631.50781.15190.40681.29491.0298-0.00012.09241.4949
RMSE0.187929.37275.47290.357525.16955.77840.028242.106812.3474
Var.0.149427.09934.14590.192123.49254.71800.028237.728610.1124
Mean0.69261.59781.35190.50681.39491.22980.89982.19242.3949
50Bias0.17170.74801.06570.33910.87620.82890.01250.46151.2245
RMSE0.143111.08456.14720.280214.47795.10080.01186.30899.8381
Var.0.113510.52515.01150.165213.71034.41360.01166.09598.3388
Mean0.67170.83801.26570.43910.97611.02890.91250.56152.1245
100Bias0.14530.62930.91900.27360.60400.81200.00770.13590.5918
RMSE0.10388.18996.70480.20468.89757.07790.00701.57613.3296
Var.0.08277.79395.86010.12978.53256.41850.00691.55752.9795
Mean0.64530.71931.11910.37360.70411.01200.90770.23591.4918
200Bias0.09290.26900.44110.19720.34350.55710.00470.00750.2014
RMSE0.06732.18013.38100.13513.77536.39460.00450.00460.4451
Var.0.05872.10783.18640.09623.65736.08410.00440.00460.4046
Mean0.59290.35900.64110.29720.44350.75710.90470.10751.1014
300Bias0.07930.15990.24050.15580.22810.42640.00490.00270.1191
RMSE0.04900.91160.89010.09931.57875.49370.00280.00180.1716
Var.0.04270.88610.83220.07511.52675.31180.00280.00180.1574
Mean0.57940.24990.44050.25580.32810.62640.90490.10281.0191
400Bias0.06060.10910.16650.13540.24170.38920.00270.00210.0736
RMSE0.03630.49730.45530.08431.32734.31970.00210.00120.0872
Var.0.03270.48540.42760.06601.26884.16830.00210.00120.0818
Mean0.56060.19910.36650.23540.34170.58920.90270.10210.9736
500Bias0.05060.05290.14830.11310.12310.30470.00290.00060.0611
RMSE0.02950.09680.49170.06420.34193.88410.00160.00070.0624
Var.0.02690.09680.46970.05140.32683.79120.00150.00070.0586
Mean0.55060.14290.34820.21310.22310.50470.90290.10070.9612

Table 6. List of CDFs for some competitive models.

Abbr.ModelParameter/
variable range 0<x<
Reference
HL-ExpGIx=1eαx1+eαxα>0Balakrishnan23
E-ExpGIIx=1eαxβα,β>0Ahuja and Nash24
ExpGIIIx=1eαxα>0Nadarajah and Kotz25
ETr-ExpGIVx=1eα1eβxα,β>0El-Alosey26
Logis-ExpGVx=111+eαx1βα,β>0Lan and Leemis27
MO-ExpGVIx=1αex11αexα>0Salah et al.28
NH-ExpGVIIx=1e11+αxβα,β>0Nadarajah and Haghighi29
Ts-ExpGVIIIx=1+λ1eαxλ1eαx2α>0,λ1Merovci and Puka30
DUS-Exp-ExpGIXx=e1eαx1e1α>0Kumar et al.11
Alp-ExpGXx=α1eβx1α1α,β>0Mahdavi and Kundu12

Table 7. Parameter estimates and standard errors (in parenthesis) along with goodness-of-fit for the failure time of 84 windshield dataset.

ModelParameters
(Standard Errors)
Goodness-of-fit
α̂β̂θ̂λ̂-LLW*A*K-S
ML-II-Exp0.8344
(0.1142)
0.9941
(0.0025)
15.3305
(9.570)
-128.73570.13530.85370.0888
Logis-Exp0.3946
(0.0193)
3.9081
(0.3917)
--131.35910.10640.70730.0862
MO-Exp11.2200
(1.9813)
---134.24960.04790.47340.0988
Alp-Exp91.0358
(45.7226)
0.8149
(0.0607)
--135.66130.06930.69140.1191
E-Exp0.7593
(0.0764)
3.5949
(0.6139)
--141.39580.21871.73910.1215
NH-Exp0.0062
(0.0073)
44.9430
(52.8504)
--145.41000.06130.60580.2588
Ts-Exp0.5717
(0.0470)
--−0.9970
(0.0503)
146.81220.18181.50000.1855
HL-Exp0.5776
(0.0497)
---153.62000.10000.93680.2632
Exp-Exp0.5137
(0.0469)
---156.18260.11691.05770.2699
Exp2.5609
(0.2776)
---164.98770.16641.39720.3033
ETr-Exp0.5431
(-)
1.2678
(-)
--164.98770.16641.39720.3033

Table 8. Parameter estimates and standard errors (in parenthesis) along with goodness-of-fit for the service times of 63 aircraft windshield dataset.

ModelParameters
(Standard Errors)
Goodness-of-fit
α̂β̂θ̂λ̂-LLW*A*K-S
ML-II-Exp0.7013
(0.4333)
0.9260
(0.0884)
3.4208
(6.852)
-98.14970.03500.23500.0657
MO-Exp5.9950
(1.2886)
---99.10040.05530.34280.0784
Alp-Exp28.4623
(22.5320)
0.8771
(0.1036)
--100.35550.09750.59320.1060
NH-Exp0.0118
(0.0098)
26.4046
(21.4902)
--100.18460.06380.39020.1436
Ts-Exp0.6790
(0.0734)
--−0.8693
(0.1465)
102.96730.16791.02490.1426
E-Exp0.6921
(0.0942)
1.8979
(0.3402)
--103.54660.20341.23150.1438
Logis-Exp0.4922
(0.0379)
2.8590
(0.3664)
--103.58500.04880.40200.0804
HL-Exp0.6871
(0.0701)
---103.84850.12010.73020.1643
Exp-Exp0.6178
(0.0665)
---105.06720.13930.84600.1699
ETr-Exp0.6989
(7.1557)
1.1588
(22.3832)
--109.29850.18611.12640.2077
Exp2.0848
(0.2626)
---109.29850.18611.12640.2078

Table 9. Parameter estimates and standard errors (in parenthesis) along with goodness-of-fit for the breaking stress of carbon fibers dataset.

ModelParameters
(Standard Errors)
Goodness-of-fit
α̂β̂θ̂λ̂-LLW*A*K-S
ML-II-Exp1.6982
(0.3168)
0.9869
(0.0056)
0.9675
(0.4008)
-142.12010.06510.41420.0602
Logis-Exp0.3850
(0.0148)
4.5316
(0.4141)
--143.31080.06680.47230.0575
E-Exp1.0133
(0.0875)
7.7926
(1.4973)
--146.18230.22671.18610.1077
Ts-Exp0.7651
(0.0704)
--−2.0285
(0.1786)
150.0987--0.1841
MO-Exp12.0519
(1.9083)
---153.66400.06470.37910.1450
Alp-Exp106.8843
(41.7742)
0.8280
(0.0528)
--153.88850.12990.65660.1431
NH-Exp0.0058
(0.0034)
48.1147
(28.6443)
--171.48350.06910.42530.2896
H-Log-Exp0.5709
(0.0451)
---181.59410.11470.58740.2904
Exp-Exp0.5061
(0.0425)
---184.89600.12710.64820.2949
Exp2.6210
(0.2621)
---196.37090.14930.76430.3206
ETr-Exp0.5467
(-)
1.1969
(-)
--196.37090.14930.76430.3206

Table 10. Parameter estimates and standard errors (in parenthesis) along with goodness-of-fit for the fatigue life of 6061 - T6 aluminum coupons dataset.

ModelParameters
(Standard Errors)
Goodness-of-fit
α̂β̂θ̂λ̂-LLW*A*K-S
ML-II-Exp0.0786
(0.0129)
0.9954
(0.0015)
1.0426
(0.4383)
-450.42210.03130.23840.0446
Logis-Exp0.0147
(0.0004)
5.5022
(0.4889)
--450.78800.03230.24470.0442
E-Exp0.0413
(0.0033)
9.6723
(1.8416)
--463.89190.27901.68120.1162
Ts-Exp0.0232
(0.0015)
−1.2479
(0.0444)
488.3510--0.2308
MO-Exp24668.86
(-)
---5831.9084.868522.7580.9899
Alp-Exp119.9019
(44.8694)
0.0316
(0.0019)
--473.82400.10630.63800.1863
NH-Exp0.0025
(0.0003)
4.7152
(0.6756)
--500.53460.09100.62140.3567
H-Log-Exp0.0221
(0.0017)
---506.31370.11650.73630.3365
Exp-Exp0.0195
(0.0016)
---509.99450.13080.81790.3407
Exp68.31
(6.8289)
---522.43490.16461.01970.3667
ETr-Exp0.02015
(0.0209)
1.2924
(2.7465)
522.43490.16461.01990.3662

Furthermore, for a visual comparison the fitted density and distribution functions, Kaplan-Meier survival and probability-probability (PP) plots, total time on test transform (TTT), and box plots, are presented in Figures 36 (a, b, c, d, e, f, g, h), respectively. These plots provide sufficient information about the closest fit to the data. All the numerical results in the subsequent tables are calculated with the assistance of statistical software RStudio-1.2.5033. with its package AdequacyModel. The explored datasets are given in the underlying data.

6d110898-0422-498a-a5a4-5ef5ca29e74e_figure3.gif

Figure 3. Failure time of 84 windshields dataset.

6d110898-0422-498a-a5a4-5ef5ca29e74e_figure4.gif

Figure 4. Service time of 63 aircraft windshield dataset.

6d110898-0422-498a-a5a4-5ef5ca29e74e_figure5.gif

Figure 5. Breaking stress of carbon fibers dataset.

6d110898-0422-498a-a5a4-5ef5ca29e74e_figure6.gif

Figure 6. Fatigue life of 6061 - T6 aluminum coupons dataset.

Conclusion

In this article, we introduced and studied a more flexible G class, called the modified Lehmann Type–II (ML–II) G class of distributions along with explicit expressions for the moments, quantile function, and OS. The exponential distribution was used as the baseline distribution for ML–II–G class, known as ML–II–Exp distribution. It was discussed comprehensively, which demonstrated the reversed-J, constant, unimodal, and right-skewed shapes of a density function. The method of MLE along with the simulation was carried out to investigate the performance of the proposed method. The efficiency of the ML–II–G class was evaluated when the most efficient and consistent results of ML–II–Exp distribution competed the well-known models and explored the dominance along with a better fit in four real-life datasets.

We hope that in the future, the proposed class and its sub-models will explore the wider range of applications in diverse areas of applied research and will be considered as a choice against the baseline models.

Data availability

Figshare. four_dataset.csv. DOI: https://doi.org/10.6084/m9.figshare.14518383.v131

This project contains the following extended data:

  • Failure Time of 84 Windshield, Service Times of 63 Aircraft Windshield, Breaking Stress of Carbon Fibers and Fatigue Life of 6061 - T6 Aluminum Coupons used for the article titled “A New Modified Lehmann Type – II G Class of Distributions: Exponential Distribution with Theory, Simulation, and Applications to Engineering Sector”

  • Ramos et al. dataset

  • Tahir et al. dataset

  • Nicholas and Padgett dataset

  • Birnbaum and Saunders dataset

Data are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC BY 4.0 Public domain dedication).

Comments on this article Comments (0)

Version 1
VERSION 1 PUBLISHED 17 Jun 2021
Comment
Author details Author details
Competing interests
Grant information
Copyright
Download
 
Export To
metrics
Views Downloads
F1000Research - -
PubMed Central
Data from PMC are received and updated monthly.
- -
Citations
CITE
how to cite this article
Balogun OS, Arshad MZ, Iqbal MZ and Ghamkhar M. A new modified Lehmann type – II G class of distributions: exponential distribution with theory, simulation, and applications to engineering sector [version 1; peer review: 1 approved with reservations, 1 not approved]. F1000Research 2021, 10:483 (https://doi.org/10.12688/f1000research.52494.1)
NOTE: If applicable, it is important to ensure the information in square brackets after the title is included in all citations of this article.
track
receive updates on this article
Track an article to receive email alerts on any updates to this article.

Open Peer Review

Current Reviewer Status: ?
Key to Reviewer Statuses VIEW
ApprovedThe paper is scientifically sound in its current form and only minor, if any, improvements are suggested
Approved with reservations A number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.
Not approvedFundamental flaws in the paper seriously undermine the findings and conclusions
Version 1
VERSION 1
PUBLISHED 17 Jun 2021
Views
18
Cite
Reviewer Report 22 Feb 2022
Nofiu Idowu Badmus, Department of Mathematics, University of Lagos, Akoka, Nigeria 
Approved with Reservations
VIEWS 18
My comments on the manuscript are as follows:

Generally, the paper is on non-normal data (i.e skewed data) as one of the challenges in Statistics in which several authors in literature have developed many suitable distributions/models to ... Continue reading
CITE
CITE
HOW TO CITE THIS REPORT
Badmus NI. Reviewer Report For: A new modified Lehmann type – II G class of distributions: exponential distribution with theory, simulation, and applications to engineering sector [version 1; peer review: 1 approved with reservations, 1 not approved]. F1000Research 2021, 10:483 (https://doi.org/10.5256/f1000research.55779.r120000)
NOTE: it is important to ensure the information in square brackets after the title is included in all citations of this article.
Views
22
Cite
Reviewer Report 22 Dec 2021
Frank Gomes-Silva, Postgraduate Program of Biometrics and Applied Statistics, Federal Rural University of Pernambuco, Recife, Brazil 
Not Approved
VIEWS 22
The authors propose a new probabilistic class. Then, starting from this class, they propose a new distribution called ML-II-Exp. Present mathematical properties, simulation, and application for this new distribution. Here are some notes:
  • The introduction can
... Continue reading
CITE
CITE
HOW TO CITE THIS REPORT
Gomes-Silva F. Reviewer Report For: A new modified Lehmann type – II G class of distributions: exponential distribution with theory, simulation, and applications to engineering sector [version 1; peer review: 1 approved with reservations, 1 not approved]. F1000Research 2021, 10:483 (https://doi.org/10.5256/f1000research.55779.r102485)
NOTE: it is important to ensure the information in square brackets after the title is included in all citations of this article.

Comments on this article Comments (0)

Version 1
VERSION 1 PUBLISHED 17 Jun 2021
Comment
Alongside their report, reviewers assign a status to the article:
Approved - the paper is scientifically sound in its current form and only minor, if any, improvements are suggested
Approved with reservations - A number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.
Not approved - fundamental flaws in the paper seriously undermine the findings and conclusions
Sign In
If you've forgotten your password, please enter your email address below and we'll send you instructions on how to reset your password.

The email address should be the one you originally registered with F1000.

Email address not valid, please try again

You registered with F1000 via Google, so we cannot reset your password.

To sign in, please click here.

If you still need help with your Google account password, please click here.

You registered with F1000 via Facebook, so we cannot reset your password.

To sign in, please click here.

If you still need help with your Facebook account password, please click here.

Code not correct, please try again
Email us for further assistance.
Server error, please try again.