Nepal Journal of Mathematical Sciences (NJMS) Research Article

ISSN: 2738-9928 (online), 2738-9812 (print) Received Date: December 7, 2024

Vol. 5, No. 2, 2024 (August): 27-38 Accepted Date: March 2, 2025
DOI: https://doi.org/10.3126/njmathsci.v5i2.76632 Published Date: March 16, 2025
School of Mathematical Sciences,
Tribhuvan University, Kathmandu, Nepal

Extended Modified Generalized Exponential

Distribution: Properties and Applications
*
Arun Kumar Chaudhary1 & Lal Babu Sah Telee2
1,2
Department of Management Science, Nepal Commerce Campus, Tribhuvan University, Nepal.
*
Corresponding Author: [email protected]

Abstract: This study presents the Extended Modified Generalized Exponential (EMGE) distribution,
a novel four-parameter model designed to improve flexibility in modeling diverse hazard rate functions,
including bathtub-shaped curves. The EMGE model is formulated by introducing an extra shape
parameter into the Modified Generalized Exponential (MGE) model, improving its capacity to
represent different failure rate patterns over time. We investigate the suggested distribution's hazard,
survival, probability density, and reversed hazard functions, among other statistical characteristics.
The EMGE model's parameters are estimated using three distinct techniques: Cramer-Von-Mises
Estimation (CVME), Least Squares Estimation (LSE), and Maximum Likelihood Estimation (MLE),
ensuring accurate and reliable parameter estimation. The performance of the model is tested on real-
world data from COVID-19 patient mortality rates, showing a strong fit to the data. Comparative
analysis with other established distributions, such as the Odd Lomax Exponential (OLE) and Modified
Weibull (MW), highlights the superiority of the EMGE model in terms of fit and information criteria
values. Our results demonstrate the potential of the EMGE distribution for improved modeling in
reliability analysis and survival data.

Keywords: Exponential distribution, Hazard rate function, Generalized exponential distribution,

Maximum likelihood function.
1. Introduction
Probability models play a crucial role in reliability analysis across diverse fields such as biological
sciences, applied statistics, and engineering. However, traditional probability models often struggle to
adequately fit reliability data, prompting researchers to modify these models for better applicability.
These modifications involve introducing additional parameters into the baseline distribution, creating
new probability models that provide a closer fit to the data compared to conventional approaches. The
incorporation of extra parameters enhances the flexibility of these models, enabling them to
accommodate a broader range of data patterns and capture complex underlying relationships more
effectively. Such improvements are particularly valuable in reliability analysis, where accurate
modeling is essential for making predictions, estimating failure rates, and assessing system
performance.
These enhanced models have practical applications in various domains, including survival analysis, risk

27
A.K.Chaudhary & L.B. Sah Telee / Extended Modified Generalized Exponential Distribution: Properties ….

assessment, and quality control. For instance, in biological sciences, they are used to model organism
lifespans or the time until disease onset, while in engineering, they help predict the reliability of
components under different operating conditions. Achieving a better fit to real-world data reduces errors
and provides more reliable insights for informed decision-making. Researchers often validate the
models through simulation studies and real-world data applications, ensuring their theoretical soundness
and computational efficiency. These advancements significantly contribute to addressing complex
challenges in reliability analysis and advancing statistical methodologies.

Generalized Exponential Distribution (GED) proposed by Gupta & Kundu [9], introduces an additional
parameter to the baseline model, enhancing its ability to represent real-world data. This added parameter
improves flexibility, enabling the GED to handle varying hazard rates, rather than the constant hazard
rate assumed by the standard exponential distribution. Chaudhary & Kumar [5] Half Cauchy modified
Exponential distribution modifying exponential model. Although effective for increasing or decreasing
hazard rate functions depending on the shape parameter, the GED cannot model complex patterns like
unimodal or bathtub-shaped hazard functions, leading to further advancements in probability modeling.
Barreto-Souza et al. [2] developed a new statistical model referred to as the beta generalized exponential
distribution. Mahmoudi & Jafari [12] recommended Generalized exponential–power series
distributions, capable of representing hazard rate functions that are increasing, decreasing, or bathtub-
shaped.
The diversity of these models is further enhanced by the Exponentiated Weibull Inverted Exponential
Model by Chaudhary et al. [7], and the Half Logistic Exponential Extension Model by Chaudhary &
Kumar, [4] as well as the Inverse Exponentiated Odd Lomax Exponential Distribution by Chaudhary et
al. [6].
These extended modifications of the exponential distribution have been developed to enhance flexibility
in modeling various types of data, particularly in reliability analysis, survival studies, and other
statistical applications. Each modification introduces additional parameters or structural changes to
better capture different hazard rate behaviors, including increasing, decreasing, bathtub-shaped, and
unimodal patterns.
Hazard rate functions (HRFs) in lifetime models often exhibit a bathtub-shaped curve, a trait commonly
observed in numerous real-world data sets. To address diverse patterns in survival and reliability
analysis, various adaptations of the Weibull distribution have been developed to improve its flexibility
and applicability. In literature, we find modifications of the original Weibull distribution to better
capture complex behaviors in data.
The following is an expression for the two-parameter Weibull distribution:
F ( y, ,  )  exp[(, y)] (1)
The earlier mentioned model does not possess a failure rate function (HRF) with a bathtub shape. To
overcome this drawback, it has been adapted into various versions that demonstrate a bathtub-shaped
hazard rate. One such modification involves utilizing the exponentiated Weibull distribution, as
proposed by Mudholkar & Srivastava [14].
Moreover, Lai et al. [10] demonstrate how adding certain constraints to the beta-integrated distribution
facilitates the design of novel lifespan distributions. These constraints further enable the derivation of
the following novel lifespan distributions.

28
 Nepal Journal of Mathematical Sciences (NJMS), Vol. 5, No. 2 , 2024 (August): 27-38

F ( y)  exp[ayb .exp( y)] (2)

The Modified Generalized Exponential distribution recommended by Telee & Kumar [19] is an
improved probabilistic model that builds upon the Generalized Exponential distribution. This
enhancement was made by introducing an additional shape parameter, increasing the model’s flexibility
and applicability in statistical analysis. The original Generalized Exponential distribution was first
developed by Gupta & Kundu [8].
The cumulative distribution function (CDF) for the Generalized Exponential distribution is given as
follows:

FGED ( x,  ,  )  (1  e x ) ; x  0,   0,   0 (3)

The Cumulative Distribution Function (or CDF) of the Modified Generalized Exponential (MGE)
model given by Telee & Kumar [19], is as follows:

 
G ( x;  ,  ,  )  1  exp  x e  x 
 ;   0,   0,   0, x  0 (4)

Probability model introduced in this study, addresses the limitations of existing models by incorporating
an additional shape parameter. This modification enhances flexibility in modeling diverse hazard rate
functions, including bathtub-shaped curves. Unlike previous distributions such as the Weibull
Extension and Modified Weibull, the EMGE model provides a better fit for real-world reliability and
survival data, as demonstrated in our comparative analysis and empirical application to COVID-19
mortality rates

Study also introduces a novel probability distribution characterized by its cumulative distribution
function (CDF) and probability density function (PDF), derived through a unique mathematical
framework. The proposed model extends traditional distributions by introducing an innovative
exponentiation mechanism that allows for greater flexibility in modeling skewed and heavy-tailed data.
This work builds upon existing probability models, particularly the Generalized Exponential and
Modified Weibull distributions, by introducing a new shape parameter. The EMGE distribution unifies
and extends existing distributions, making it applicable to a broader range of real-world datasets,
including those with non-monotonic failure rates.
2. Extended Modified Generalized Exponential (EMGE)Distribution

The Extended Modified Generalized Exponential Distribution is created by adding an additional shape
parameter to the Modified Generalized Exponential Distribution's cumulative distribution function (or
CDF), as created by Telee & Kumar [19] in equation (4). The proposed Extended Modified Generalized
Exponential (EMGE) model is characterized by its distribution and density functions, which are defined
as follows:

   e   x 
F ( x;  ,  ,  , )  1  1  exp    ; x,  ,  ,  ,   0 (5)
  x 
 

 1
 e  x    e   x    e   x 
f ( x;  ,  ,  ,  )   
 x



 
    x 1 exp 
 x

 1  exp 


 x




; x,  ,  ,  ,   0 (6)

29
A.K.Chaudhary & L.B. Sah Telee / Extended Modified Generalized Exponential Distribution: Properties ….

2.1 Survival Function

The reliability function associated with the proposed model is defined in equation (7).

   e   x 
S ( x;  ,  ,  , )  1  exp    ; x,  ,  ,  ,   0 (7)
  x 
 

2.2 Hazard Rate Function

Equation (8) provides a mathematical expression for the failure rate function, which represents
the evolving failure probability over time.
1
 e  x    e   x    e   x 
h( x)   
 x



 
    x 1 exp 
 x

 1  exp 


 x




(8)

2.3 The Reversed hazard function (RHR)

Equation (9) represents the reversed hazard rate function.

 1    1
 e  x    e   x    e   x  
1  1  exp   e
 x 
RHR( x)   
 x



 
    x 1 exp 
 x

 1  exp 


 x




   x





 (9)
  

2.4 Cumulative hazard function (CHF)

The CHF for the recommended model is shown in equation (10).

   e   x 
CHF ( x)   log[ S ( x)]   log 1  exp    (10)
  x 

2.5 The Quantile function

Equation (11) specifies the quantile function for the EMGE, based on the assumption that u is
uniformly distributed over the interval [0,1].

1  x ln[1  (1  u )1/ ] 
Q u    ln   (11)
   

2.6 Skewness and Kurtosis

The following formula can be used to get the quartile-based coefficient of skewness.
Q3  Q1  2Q2
Sk  (12)
Q3  Q1

Moors [13] states that the following formula may be used to define the kurtosis coefficient
based on octiles:

30
 Nepal Journal of Mathematical Sciences (NJMS), Vol. 5, No. 2 , 2024 (August): 27-38

Q  0.375  Q  0.875  Q  0.125  Q  0.625

KM  (13)
Q  0.75   Q  0.25 
“

Figure 1 displays the hazard rate function (HRF) and probability density function (PDF) of the
suggested model at constant lambda = 3 & theta = 0.5. The PDF demonstrates a unimodal
distribution with positive skewness, indicating that most values cluster near the center. In
contrast, the HRF has an inverted bathtub shape and an increasing pattern, suggesting a variety
of risk patterns over time.

Figure 1: Density function (left) and hazard function (right) for lambda = 3 & theta = 0.5

3. Estimation Methods

MLE, LSE, and CVM methods are three well-known techniques used to estimate the
parameters of the suggested EMGE model.
Maximum Likelihood Estimation (MLE)
A common statistical technique for estimating distribution parameters is Maximum Likelihood
Estimation (MLE), which maximizes the likelihood function. In the case of the EMGE model,
the likelihood function is formulated using its PDF, and the parameters are estimated by
maximizing this function with respect to the unknown parameters.
Log likelihood function for EMGE is given as
 e  xi  n n   e
  xi 1     e  xi 1 
l  n log( ) 
n

i 1
log 
 x  i 1  
   log    xi 1   
i 1  
    1  log 1  exp 

n
i 1   x 
  (14)

 i   xi 1   i 1 
Partial derivatives can be obtained by differentiating equation (14) and solving derivatives,
parameters can be estimated. Solution of these nonlinear equations are quite rigorous
analytically, so the likelihood function is maximized using the Newton-Raphson algorithm,
which aids in the development of the observed information matrix. Consequently, the variance-
covariance matrix obtained is given by equation (15)

31
A.K.Chaudhary & L.B. Sah Telee / Extended Modified Generalized Exponential Distribution: Properties ….

 V11 (ˆ ) V12 (ˆ , ˆ ) V13 (ˆ , ˆ ) V14 (ˆ ,ˆ) 

 
 
1
 V21 ( ˆ , ˆ ) V22 ( ˆ ) V23 ( ˆ , ˆ ) V24 ( ˆ ,ˆ) 
  H   | ˆ      (15)
 V31 (ˆ, ˆ ) V32 (ˆ, ˆ ) V33 (ˆ ) V34 (ˆ,ˆ) 
 ˆ ˆ ˆ ˆ ˆ ˆ 
“

 V41 ( , ˆ ) V42 ( ,  ) V43 ( ,  ) V44 ( )  ”

To construct approximately 100(1-δ) % confidence intervals for estimating α, β, λ and θ, the

following method can be utilized, leveraging the asymptotic normality property of Maximum
Likelihood Estimates (MLE).

“
ˆ  Z / 2 V11 (ˆ ) , ˆ  Z / 2 V22 ( ˆ ) , ˆ  Z / 2 V33 (ˆ ) and ˆ  Z / 2 V44 (ˆ)
”
(16)

In this context, Z /2 refers to the upper critical value of the standard normal distribution.

Least Squares Estimation (LSE)

The total of the squared differences between the observed data points and the cumulative
distribution values that the model predicts is minimized using the Least Squares Estimation
(LSE) approach. This technique is particularly useful when dealing with data where the
assumption of normality may not hold.
Cramér–von Mises (CVM) Criterion
The difference between the observed and theoretical distribution functions is reduced by
applying the CVM approach. It offers an additional reliable technique for parameter estimation
and is based on the weighted squared discrepancies between the theoretical and empirical
cumulative distribution functions. These estimation techniques provide a comprehensive
approach to fitting the EMGE model to data, ensuring that the parameters can be estimated
with high precision and suitability for various applications in statistical modeling. By utilizing
these methods, we aim to provide efficient and reliable estimates for the parameters of the
EMGE distribution, enhancing its applicability in diverse fields such as reliability analysis,
survival analysis, and other domains where flexible and accurate modeling of data is crucial.
4. Application to Real Dataset
According to Bantan et al. [1], the suggested model was tested on a real-world dataset that comprised
the COVID-19 pandemic mortality rates in Mexico from March 4, 2020, to July 20, 2020, for 106
patients. The rates were divided by five to make them easier to analyze. The following is how the dataset
is displayed:
1.7652, 1.2210, 1.8782, 2.9942, 2.0766, 1.4534, 2.6440, 3.2996, 2.3330, 1.2030, 2.1710, 1.2244,
1.3312, 0.6880, 1.1708, 2.1370, 2.0070, 1.0484, 0.8668, 1.0286, 1.5260, 2.9208, 1.5806, 1.2740,
0.7074, 1.2654, 0.9460, 0.6430, 1.8568, 2.5756, 1.7626, 2.0086, 1.4520, 1.1970, 1.2824, 0.6790,
0.8848, 1.9870, 1.5680, 1.9100, 0.6998, 0.7502, 1.3936, 0.6572, 2.0316, 1.6216, 1.3394, 1.4302,
1.3120, 0.4154, 0.7556, 0.5976, 0.6672, 1.3628, 1.5708, 1.6650, 1.7120, 0.6456, 1.4972, 1.3250,
1.2280, 0.9818, 0.9322, 1.0784, 2.4084, 1.7392, 0.3630, 0.6654,1.0812, 1.2364, 0.2082, 0.3600,
0.9898, 0.8178, 0.6718, 0.4140, 0.6596, 1.0634, 1.0884, 0.9114, 0.8584, 0.5000, 1.3070, 0.9296,
0.9394, 1.0918, 0.8240, 0.7884, 0.6438, 0.2804, 0.4876, 0.6514, 0.7264, 0.6466, 0.6054, 0.4704,
0.2410, 0.6436, 0.5852, 0.5202, 0.4130, 0.6058, 0.4116, 0.4652, 0.5052, 0.3846

32
 Nepal Journal of Mathematical Sciences (NJMS), Vol. 5, No. 2 , 2024 (August): 27-38

The boxplot and TTT plot for the examined dataset are shown in Figure 2. A boxplot with a positive
skew shows that the data is not distributed normally. Likewise, the concave shape of the TTT curve
signifies an increasing failure rate.

Figure2: Boxplot (Left panel) and TTT plot (Right panel) of the data
The dataset's descriptive statistics, shown in Table 1, show that it is positively skewed and
deviates from normalcy.
Table 1: The data's descriptive statistics

“ Min Q1 Md X̅ Q3 Sd Skewness Kurtosis Max ”

0.2082
“ 0.66 1.06 1.165 1.52 0.65 0.973 3.67 3.2996 ”

The MLE, LSE, and CVME methods are used to estimate the model's parameters using the R
software's optim() function (R Core Team[16]).The calculated parameters and the associated
standard errors (SE) are shown in Table 2.

Table 2: Estimated Parameters using MLE, LSE and CVME along with respective S.E.

Methods Alpha Beta Lambda Theta

MLE 7.1294 0.2362 3.0860 0.5491

LSE 0.6660 3.3078 1.4800 1.9500

CVME 0.2182 0.0002 0.0002 4.9790

Furthermore, for each of the three estimating methods, the Log-Likelihood values and several
information criteria, including BIC, AIC, CAIC, and HQIC, were computed in table 3.

33
A.K.Chaudhary & L.B. Sah Telee / Extended Modified Generalized Exponential Distribution: Properties ….

Table 3: The BIC, HQIC, AIC, CAIC, and log likelihood (LL)
Model LL AIC BIC CAIC HQIC
MLE -90.70515 189.4103 200.0641 189.8063 193.7283

LSE -94.75668 197.5134 208.1671 197.9094 201.8314

CVM -96.37799 200.756 211.4097 201.1520 205.074

The goodness-of-fit test results are summarized in Table 4, which provides the statistics for
Anderson-Darling (A²), Cramér-von Mises (W), and Kolmogorov-Smirnov (KS), along with
their respective p-values for different estimation approaches. Additionally, Table 4 compares
the performance of various estimation techniques by evaluating how well they fit the observed
data.
Table 4: Statistics for KS, W, and A2 together with associated p-values
“ ”

Methods KS(p-value)
“ W(p-value) A2(p-value) ”

“ MLE 0.0817(0.4790) 0.1051(0.5613) 0.7129(0.5477)

LSE 0.0483(0.9652) 0.0347(0.9589) 0.2302(0.9798)
CVME 0.0486(0.9637) 0.0327(0.9669) 0.2049(0.9890) ”

Plots of PDF and CDF are frequently used to evaluate how well a given model fits data.
Additionally, generating Q-Q and P-P plots provides deeper insights into the model's
performance. The P-P plot highlights areas where the model may not fit well, while the Q-Q
plot emphasizes the alignment in the tails of the distribution. Figure 3 demonstrates the EMGE
model's strong fit to the data. The EMGE model's goodness of fit is assessed in a thorough
manner by integrating Q-Q and P-P plots with PDF and CDF plots.

Figure 3: The EMGE model's P-P (left) and Q-Q (right) graphs.

34
 Nepal Journal of Mathematical Sciences (NJMS), Vol. 5, No. 2 , 2024 (August): 27-38

The empirical cumulative distribution function (ECDF), density plot, and histogram are
compared with the fitted CDF in Figure 4.

Figure 4: Ecdf against fitted cdf (right) and histogram versus pdf plot (left).

For model comparison, five previously published probability models were evaluated. Telee and Kumar
[18] created the Lindley Generalized Inverted Exponential (LGIE) model, Lai et al. [11] developed the
Modified Weibull (MW) model, Chaudhary and Kumar [3] developed the Logistic Inverse Exponential
(LIE) distribution, Tang et al. [17] described the Weibull Extension (WE) model, and Ogunsanya et al.
[15] introduced the Odd Lomax Exponential (OLE) distribution.

Table 5 provides the estimated parameter values and their corresponding standard errors for the
proposed model in comparison with other models. These estimates are essential for evaluating and
benchmarking the performance of the models. Furthermore, Table 6 showcases important statistical
metrics, shedding light on the precision and reliability of these parameter estimates. Assessing the
robustness and effectiveness of the proposed model in capturing the underlying relationships in the data
in comparison to other models is made easy by this comprehensive analysis.

Table 5: “Values of estimated parameters for EMGE & their SE, along with competing models”
Model Alpha Beta Theta Lambda”
EMGE 7.1293 0.2361 0.5491 3.0860
OLE 0.1479 0.0119 - 0.1059
LGIE 7.7120 - 0.6487 1.4727
WE 20.2560 1.9589 - 10.3291
MW 0.5718 1.8937 - 0.0225
LIE 2.0429 - - 0.6717

The effectiveness of the proposed model is assessed using various criteria, including the Bayesian
Information Criterion (BIC), Hannan-Quinn Information Criterion (HQIC), Corrected Akaike
Information Criterion (CAIC), and Akaike Information Criterion (AIC).A summary of the results
obtained is provided in Table 6.

35
A.K.Chaudhary & L.B. Sah Telee / Extended Modified Generalized Exponential Distribution: Properties ….

Table 6: The HQIC, AIC, CAIC, BIC, and log likelihood (LL)
Model LL AIC BIC CAIC HQIC
EMGE -90.71 189.41 200.06 189.81 193.73
“OLE -92.49 190.98 198.97 191.21 194.22
LGIE -93.152 192.30 200.29 192.54 195.54
WE -93.79 193.58 201.57 193.81 196.82
MW -93.86 193.71 201.70 193.95 196.95
LIE -96.39 196.78 202.10 196.89 198.94”

Simulation Study
Here, we have performed simulation analysis for the EMGE model. It helped in understanding its
estimation properties, evaluating the performance of statistical methods, and determining its
applicability to different types of data. It provides a foundation for making informed decisions about
the model's effectiveness and its potential improvements for practical use in various domains. Below is
a table summarizing the bias and mean squared error (MSE) for different combinations of sample size
(n) and number of simulations (k) at α = 1.2, β = 1.0, λ = 2.0 and θ = 1.2:
Table 7: Bias and MSE for α = 1.2, β = 1.0, λ = 2.0 and θ = 1.2.
Sample Size (n) Simulations (k) Bias MSE
50 500 0.0012 0.0044
50 1000 0.0061 0.0042
50 1500 0.0082 0.0042
100 500 0.0050 0.0023
100 1000 0.0071 0.0022
100 1500 0.0066 0.0023
150 500 0.0039 0.0014
150 1000 0.0052 0.0015
150 1500 0.0047 0.0015
200 500 0.0045 0.0011
200 1000 0.0064 0.0012
200 1500 0.0048 0.0011
250 500 0.0064 0.0009
250 1000 0.0064 0.0009
250 1500 0.0059 0.0009
300 500 0.0057 0.0007
300 1000 0.0072 0.0008
300 1500 0.0048 0.0008
350 500 0.0071 0.0007
350 1000 0.0066 0.0007
350 1500 0.0065 0.0007
400 500 0.0061 0.0006
400 1000 0.0061 0.0006
400 1500 0.0060 0.0006
The simulation study demonstrates that larger sample sizes (n) lead to more reliable and precise
estimates, as reflected in the decreasing bias and MS. MSE consistently decreases with increasing n and
k reinforcing the understanding that larger sample sizes and a higher number of simulations lead to
better estimations of the true parameter values.

36
 Nepal Journal of Mathematical Sciences (NJMS), Vol. 5, No. 2 , 2024 (August): 27-38

5. Conclusion

This research proposes the Extended Modified Generalized Exponential (EMGE) distribution as a
valuable framework for reliability assessment and survival analysis. The EMGE model incorporates
intricate patterns in hazard rates, such as those with bathtub-shaped curves, by adding a shape parameter
to the Modified Generalized Exponential distribution. The comparative analysis with alternative models
confirms the EMGE’s superior performance in terms of fit and statistical criteria.

The derived CDF and PDF exhibit improved adaptability in representing diverse datasets, which is
crucial for applications in reliability analysis, survival modeling, or econometrics. The mathematical
derivation is rigorously validated, ensuring consistency with fundamental probability properties. Also,
the model provides a better fit compared to existing distributions, as demonstrated through empirical or
simulation-based comparisons. Furthermore, this model's ability to handle diverse failure rate functions
makes it a valuable addition to the toolkit for reliability analysis, offering more accurate insights into
system behavior and improving decision-making processes in various fields. Future work can explore
further applications of the EMGE distribution in different domains, as well as potential extensions to
address even more complex data patterns. Also, proposed distribution introduces additional parameters
and a unique exponentiation mechanism, which provides more flexibility compared to traditional
distributions.

References

[1] Bantan, R. A., Ahmad, Z., Khan, F., Elgarhy, M., Almaspoor, Z., Hamedani, G. G., & Gemeay,
A. M. (2023). Predictive modeling of the COVID-19 data using a new version of the flexible
Weibull model and machine learning techniques. Mathematical Biosciences and Engineering. Eng,
20(2): 2847-2873. http://dx.doi.org/10.3934/mbe.2023134.
[2] Barreto-Souza, W., Santos, A. H., & Cordeiro, G. M. (2010). The beta generalized exponential
distribution. Journal of Statistical Computation and Simulation, 80(2):159-172.
https://doi.org/10.1080/00949650802552402.

[3] Chaudhary, A. K., & Kumar, V. (2020a). Logistic inverse exponential distribution with properties
and applications. International Journal of Mathematics Trends and Technology-IJMTT, 66(10).
https://doi.org/10.14445/22315373/IJMTT-V66I10P518.

[4] Chaudhary, A. K., & Kumar, V. (2020b). Half logistic exponential extension distribution with
Properties and Applications. International Journal of Recent Technology and Engineering
(IJRTE), 8(3): 506-512. http://www.doi.org/10.35940/ijrte.C4625.099320.

[5] Chaudhary, A. K., & Kumar, V. (2022). Half Cauchy-modified exponential distribution: properties
and applications. Nepal Journal of Mathematical Sciences, 3(1): 47-58.

37
 A.K.Chaudhary & L.B. Sah Telee / Extended Modified Generalized Exponential Distribution: Properties ….

[6] Chaudhary, A. K., Telee, L. B. S., & Kumar, V. (2022). Inverse exponentiated odd Lomax
exponential distribution: Properties and applications. Nepalese Journal of Statistics, 29-50.
https://doi.org/10.3126/njs.v6i01.50801.

[7] Chaudhary, A. K., Telee, L. B. S., & Yadav, D. K. (2021). Exponentiated Weibull inverted
exponential distribution: Model, properties and applications. NCC Journal, 6(1): 1-10.
https://doi.org/10.3126/nccj.v6i1.57785.
[8] Gupta, R. D., & Kundu, D. (1999a). Theory & methods: Generalized exponential distributions.
Australian & New Zealand Journal of Statistics, 41(2): 173-188. https://doi.org/10.1111/1467-
842X.00072.
[9] Gupta, R. D., & Kundu, D. (2007). Generalized exponential distribution: Existing results and some
recent developments. Journal of Statistical Planning and Inference, 137(11): 3537-3547.
https://doi.org/10.1016/j.jspi.2007.03.030.
[10] Lai, C. D., Jones, G., & Xie, M. (2016). Integrated beta model for bathtub-shaped hazard rate data.
Quality Technology & Quantitative Management, 13(3): 229-240.
https://doi.org/10.1080/16843703.2016.1180028.
[11] Lai, C. D., Xie, M., & Murthy, D. N. P. (2003). A modified Weibull distribution. IEEE
Transactions on Reliability, 52(1): 33-37. https://doi.org/10.1109/TR.2002.805788.
[12] Mahmoudi, E., & Jafari, A. A. (2012). Generalized exponential–power series distributions.
Computational Statistics & Data Analysis, 56(12): 4047-4066.
https://doi.org/10.1016/j.csda.2012.04.009.
[13] Moors, J. J. A. (1988). A quantile alternative for kurtosis. Journal of the Royal Statistical Society:
Series D (The Statistician), 37 (1): 25-32. https://doi.org/10.2307/2348376
[14] Mudholkar, G. S., & Srivastava, D. K. (1993). Exponentiated Weibull family for analyzing bathtub
failure-rate data. IEEE Transactions on Reliability, 42(2): 299-302.
https://doi.org/10.1109/24.229504.
[15] Ogunsanya, A. S., Sanni, O. O., & Yahya, W. B. (2019). Exploring some properties of odd Lomax-
exponential distribution. Annals of Statistical Theory and Applications (ASTA), 1: 21-30.
[16] R Core Team (2023). R: A Language and environment for statistical computing. (Version 4.1)
[Computer software]. Retrieved from https://cran.r-project.org. (R packages retrieved from CRAN
snapshot 2023-04-07).

[17] Tang, Y., Xie, M., & Goh, T. N. (2003). Statistical analysis of a Weibull extension model.
Communications in Statistics-Theory and Methods, 32(5): 913-928.
[18] Telee, L. B. S., & Kumar, V. (2021). Lindley generalized inverted exponential distribution: model
and applications. Pravaha, 27(1): 61-72. https://doi.org/10.3126/pravaha.v27i1.50616

[19] Telee, L. B. S., & Kumar, V. (2023). Modified generalized exponential distribution. Nepal Journal
of Mathematical Sciences, 4(1):21-32. https://doi.org/10.3126/njmathsci.v4i1.53154.



38