Fuzzy Reliability Analysis of the COVID-19 Mortality Rate Using a New Modified Kies Kumaraswamy Model

Riad, Fathy H.; Alruwaili, Bader; Almetwally, Ehab M.; Hussam, Eslam

doi:https://doi.org/10.1155/2022/3427521

Journal of Mathematics

On this page

Abstract Introduction Additional Points Data Availability Conflicts of Interest References Copyright Related Articles

Special Issue

Analytical Methods to Model Nature

View this Special Issue

Research Article | Open Access

Volume 2022 | Article ID 3427521 | https://doi.org/10.1155/2022/3427521

Fuzzy Reliability Analysis of the COVID-19 Mortality Rate Using a New Modified Kies Kumaraswamy Model

Fathy H. Riad,^1,2Bader Alruwaili,¹Ehab M. Almetwally,^3,4and Eslam Hussam⁵

Academic Editor: Nauman Raza

Received06 Jan 2022

Revised24 Feb 2022

Accepted07 Mar 2022

Published29 Apr 2022

Abstract

In this paper, we developed a novel superior distribution, demonstrated and derived its mathematical features, and assessed its fuzzy reliability function. The novel distribution has numerous advantages, including the fact that its CDf and PDf have a closed shape, making it particularly relevant in many domains of data science. We used both conventional and Bayesian approaches to make various sorts of estimations. A simulation research was carried out to investigate the performance of the classical and Bayesian estimators. Finally, we fitted a COVID-19 mortality real data set to the suggested distribution in order to compare its efficiency to that of its rivals.

1. Introduction

Nowadays we face a huge waves of viruses such as COVID-19, which attracted a great attention in the last two years. However, as statisticians we must play an important role in understanding and modelling the COVID-19 infections, so we had to make a statistical model that is capable of fitting and modelling them, whether it is continuous or discrete random variable. In this paper, we made a lot of effort to choose an appropriate model which is superior in fitting the COVID-19 infections in many countries. At last, we settled to introduce a new superior distribution as mixture of other ones to overcome the defects of the baseline distribution.

Over the years and decades, modelling for lifetime distributions has garnered considerable attention, and their popularity has increased as the relevance of modelling phenomena and pandemics has grown. Researchers in distribution theory choose to model the data whether by adding a new parameter that increases the flexibility of the distribution of interest or by creating a new distribution family. Modelling is extremely useful in a wide variety of sectors, including commerce, economics, dependability, and health research. See [1] for more reading. In the next paragraph we will discuss in details the new proposed distribution.

We developed a statistical model for COVID-19 mortality data in Saudi Arabia in this study; for further information about the other models, refer to Kumar [2], Khakharia et al. [3], Shafiq et al. [4], Sindhu et al. [5], Wang [6], Lalmuanawma et al. [7], and Bullock et al. [8].

The new proposed model is a combination between the Kumaraswamy distribution’s (Ku) and the modified Kies distribution. The CDF of Ku distribution and PDF with parameters were put forth by [9] in the following manner:

Many authors worked on extensions of the Kies distribution as an example; see Kumar and Dharmaja [10] who worked on the exponentiated Kies distribution. Also, Dey et al. [11] worked on the modified Kies distribution when the sample they worked on censored data specially under type II progressive censored sample. Al-Babtain et al. [12] developed a different distribution family based on the modified Kies (MK) family. Let us say that is the baseline CDF depending on a parameter vector ; (3) defines the CDF of the MK family. Kumar and Dharmaja examined the characteristics of the exponentiated Kies distribution as we can see in reference [10]. One of the authors who have created a MK distribution for a type II progressive censored sample; in addition, the estimation of the distribution parameters has been obtained (see Dey et al. [11] for more information). Referring to Al-Babtainet al. [12], if is the baseline CDF , then is the MK family’s CDF that may be expressed as follows:where is the parameter vector and is parameter vector of .

The PDF of (3) is as follows:

Using the distributions above, the two-parameter modified Kies Kumaraswamy (MKKu) is produced in this paper. The MKKu has a number of interesting properties. The suggested MKKu distribution has an extremely adaptable PDF since it may be positive skewed, negative skewed, or symmetric, which allows for more flexibility for tail. It is capable of simulating declining, increasing, bathtub, and reverse-J hazard rates. Also, the suggested distribution has a precise closed-form CDF and is extremely easy to manage. These advantages make the distribution an attractive choice for applications in a variety of fields, including biological life testing, dependability, and actuarial data. This article showed one real-world data application and demonstrated via modelling that the novel distribution is a great rival to many well-known and conventional distributions with the parameters of the scale and shape such as type II power Topp-Leone inverse exponential (Bantan et al. [13]), Topp-Leone generalized exponential (Kunjiratanachot et al. [14]), Kumaraswamy, beta, Gompertz Lomax (Oguntunde et al. [15]), alpha power inverted TL (Ibrahim et al. [16]), and MK exponential (Al-Babtain et al. [12]) distributions.

The paper is arranged and organized as follows. In Section 2, the MKKu distribution is obtained. In Section 3, some mathematical characteristics of the MKKu distribution are discussed. In Section 4, we establish an MKKu distribution estimate technique. In Section 5, a fuzzy reliability function was obtained. In Section 6, we present the simulation study for the estimation techniques and concluded our major findings in this section. In Section 7, we acquired applications of real-world data analysis. Section 8 summarises and concludes the article. Future research will include the development of an extension for the bivariate MK inverted Topp-Leone based on copula as in [17]. We propose to investigate a unique application of the MKITL distribution using a censored sample; see [18] for further information.

2. MKKu Distribution

By employing the Kumaraswamy distribution with these parameters and the CDF specified (for 0 < x < 1), (1) and (2) are used. We construct the CDF for MKKu distribution by putting the Ku distribution’s CDF into (3).

The corresponding PDF iswhere is parameter vector .

The survival function of the MKKu distribution is represented by

The function of hazard rate (HR) of the MKKu distribution is given by

Figures 1 and 2 provide specific plots of the MKKu distribution for the values indicated for and in the equation of the MKKu distribution. It can be seen by the diagrams provided in Figure 2 that the MKKu distribution HR function may be increased, lowered, and shaped like a bathtub. In contrast to a Ku distribution, which is a bad model for the data and phenomena that indicate growing, decreasing forms, and failure rates of the bathtub, the MKKu distribution is more adaptable in terms of evaluating lifespan data than its rival; this is one of its advantages.

3. The Suggested Distribution Statistical Properties

3.1. A General Expansion of the MKKu Density

In this article, we provide a linear representation for the MK family and utilise it to construct a suitable linear representation for the MKKu distribution.

The last equation of the MKKu distribution may be reformulated using the PDF and CDF of the Ku distribution.

Equation (10) denotes the Ku density with parameter .

3.2. Quantile for the MKKu Distribution

In this part of the paper, we drive the formula of the MKKu distribution’s quantile function; for example, let us say that is derived by inverting (5) as follows:

The first quartile, denoted by the symbol , the second quartile, denoted by the symbol , and the third quartile, denoted by the symbol , are produced by putting U = 0.25, 05, 0.75, accordingly in (11).

3.3. The Proposed Distribution Related Moments

The r_th of the distribution can be find by using (10) as

4. Estimation Techniques for Distribution Parameters

Using Bayesian and non-Bayesian estimate approaches, we are able to solve the estimation issue for the MKKu distribution parameters. The maximum likelihood estimators (MLE), maximum product of spacings, Anderson–Darling, Cramér–von-Misse, and Bayesian approaches based on the squared error loss function (SELF) are used. These non-Bayesian methods have been discussed to estimate parameters for different models as per [19–21], while for Bayesian estimation, see [22–24].

4.1. The MLE Estimator

Suppose is a random sample of any size, then likelihood function of the proposed distribution can be written as follows:

For such MKKu distribution, the log-likelihood formula is represented by

We will differentiate with respect to every parameter in the distribution

We can evaluate the estimates of the three parameters , , and by determining the greatest value for (14) and then by computing its first derivative in terms of , , and . Using the Newton-Rapshon approach, the R program may be used to maximize the log-likelihood and yield the MLE.

4.2. Maximum Product of Spacings (MPS) Method

MPS approach was established by Cheng and Amin [25] and has been employed in censored applications as [26–28]. Assume . Given a random sample of size , the MKKu distribution’s uniform spacing may be written as indicate the uniform spacings, , and . The MPS estimators (MPSE) of the MKKu parameters can be found by maximizingwith respect to , and .

The MPSE of the MKKu parameters can found by solving

4.3. Cramer–von-Mises (CVME) Method

This method may be calculated using the difference between both the CDF and empirical distribution function estimations by Luceno [29]. The CVME values for the MKKu parameters are produced by minimizing the next function in terms of , and :

Further, the CVME follow by solving the next nonlinear equations

Solve these equations to get the estimates (20)–(22).

4.4. Anderson–Darling Method

Another sort of minimal distance estimators is the (ADE) Anderson–Darling estimators, minimizing the ADE of the MKKu parameters with respect to , and . Additionally, the ADE may be produced by solving nonlinear equations, which yields the ADE of the MKKu parameters estimates.

4.5. Bayesian Estimation Method

This section contains Bayesian estimations of , and . We assume that , and have gamma priors; also, the gamma prior density was first used. The following are the gamma priors for distribution parameters:

As we know the two priors are independent of each other, so the joint prior of and is obtained as

To get acceptable and superior values for the independent joint prior’s hyperparameters, we may employ the MLE method’s estimate and variance-covariance matrix. The calculated hyperparameters may be represented by equating the mean and variance of gamma priors:where is the number of iteration.

By multiplying (13) by (25) and making some simplifications, the posterior distribution is formed as follows:

As , the Bayesian estimations of the distribution parameter using the squared error loss function (SELF) is

It is obvious that SELF estimations of in (28) is considered as the division of more than one integration over each other. As the multiple integrals are hard to deal with, so we used Mathematica 11 to apply an approximation technique, which is very useful in solving these kinds of integration. So in this context, we used the Markov Chain Monte Carlo (MCMC) technique to get an approximate value of integrals. The important methods of the MCMC technique is the Metropolis-Hastings (MH) algorithm. The following approach is used to produce random samples from conditional posterior densities of the MKKu distribution:

5. Fuzzy Reliability

The concepts of dependability and HR function are probabilities that describe the life time (a random variable) from the start of a failure to the point where we need to adjust the manufacturing process. The use of dependability and HR functions has been broadened now that fuzzy factors have been included to these functions. The lifespan probability of the system’s components is treated as accurate numbers in classic reliability models. In the real world, however, this accuracy of system lifetimes is not true since the values of system parameters acquired via experiment, faulty measurement, human judgement, or guess are all subject to some uncertainty [30]. As a result, incorporating the idea of fuzziness will be more acceptable. As a result, while studying system behaviour or addressing system reliability, it will be more suitable to include the idea of fuzziness. The membership functions are utilised to describe the level of fuzziness associated with the lifespan data or system parameters in order to generate the fuzzy parameters [31]. The fuzzy set theory may be used to explain the real world in a realistic and practical way. Zadeh [32] was the one who introduced it.

5.1. Basic Concepts in the Theory of Fuzzy Sets

The data type utilised to estimate the parameters, as well as the trustworthiness of the probability distribution, is critical to the correctness of the findings we acquire. As a result, the data type has to be given. Fuzzy data are one of these sorts of data, and it is one of the new and essential developments in statistics since many occurrences in our actual world have no precise limits. They, too, lack precision in their measurements. As a result, we will go over some fundamental notions in the theory of fuzzy sets [31].

Let be a continuous random variable that represents a system’s failure time (component). Chen and Pham [33] may then calculate the fuzzy dependability using the fuzzy probability formula:where is a membership function that indicates the degree to which each element of a given universe belongs to the fuzzy set.

Now, assume that is

The lifespan for can be calculated using the computational approach of the function of fuzzy numbers and corresponds to a given value of as follows (Chen and Pham [33]):

then

Thus, the fuzzy reliability values can be obtained for all values of , and we consider fuzzy reliability of MKKu distribution depends on the definition of fuzzy reliability as per the following equations:

If ,

If has value,where , and if ,

6. Simulation Results

In this part, we estimated the MKKu parameters using two different approaches and compared their performance using a simulation study. We explore several sample sizes for various parameter choices of , and . We produce random samples of from the MKKu distribution. We derive the relative bias (RB) values of the estimates, the mean square error (MSE), and the accompanying fuzzy reliability for each estimate as follows:

The effectiveness of different estimators is quantified in terms of RB, MSE, and confidence interval, with the most effective methods having MSE values near to zero. The R software is used to acquire the simulation results. Tables 1 and 2 show the estimates RB, MSE, and confidence interval for MLE and Bayesian.

6.1. Concluding Remarks on the Simulation’s Findings

(1)All estimation methods work admirably and have a very tiny MSE, and their mean values trend to the original values of the parameters.(2)The discrepancies in the results of all estimators are extremely minor, both in terms of the MSE and the average of the parameters.(3)The fuzzy reliability with high value of cut has the largest value with the largest MSE.(4)In median sample size, we note the MPS is better than MLE.(5)We noted that the CVM estimation method is better than MLE, MPS, and AD estimation method in Table 2.(6)We noted that AD estimation method is better than MLE, MPS, and AD estimation method in Table 1.(7)We noted that the Bayesian estimation method is better than another estimation methods.

7. Analyses of Data Used as an Application to the Proposed Distribution

This section will demonstrate the MKKu distribution’s potential for real-world data set. The MKKu distribution is compared to many different models, including Kumaraswamy, Beta, and Gompertz Lomax (GL) (Oguntunde et al. [15]), Topp-Leone generalized exponential (TLGE) [14], Type II Power Topp-Leone invers exponential (TIIPTLIE) [13], Kumaraswamy-Kumaraswamy (KuKu) [34], alpha power inverted TL (APITL) [16], Toop-Leone Weibull Lomax (TLWL) [35], and Marshall-Olkin Kumaraswamy (MOKu) [36] distributions.

Table 3 contains the values of the AIC, BIC, CAIC, HQIC, Cramer-von Mises statistics (WS), Anderson–Darling statistics (AS), and Kolmogorov-Smirnov statistics (KSS), as well as their associated P-values, for all models fitted using a real data set. Additionally, these tables provide the mean and standard error (SE) of the parameters for the investigated models. Figure 3 illustrates the fit empirical, histogram, and PP-plot for the MKKu distribution for Saudi Arabia’s COVID-19 data.

The data contains 40 days of COVID-19 data from Saudi Arabia, from 22 July to 30 August 2021 (see the link [37]). These results illustrate the drought mortality rate. The data are as follows: 0.2375 0.2962 0.2167 0.2752 0.2353 0.2347 0.1951 0.2140 0.2329 0.2711 0.2126 0.2314 0.1924 0.2113 0.2683 0.2487 0.2674 0.1716 0.2666 0.2091 0.2278 0.1706 0.2271 0.1890 0.2077 0.2452 0.1319 0.2259 0.1504 0.1879 0.1689 0.2063 0.2249 0.1686 0.1310 0.1497 0.1309 0.1495 0.1121 0.1120. The results of application of Saudi Arabia are shown in Table 3 and Figure 3.

Table 3 shows MLE and different measures for parameter estimates of MKKu distribution for COVID-19 data of Saudi Arabia. Table 4 shows different estimators by using different estimation methods. Figure 4 shows convergence plots of MCMC for parameter estimates of MKKu distribution. Figure 5 shows density of posterior marginal with histogram of MCMC results for parameter estimates of MKKu distribution.

7.1. Concluding Remarks on the Real Data Application

When compared to all other models employed to fit the COVID-19 data, the MKKu distribution has the greatest P-value and the smallest distance of Kolmogorov-Smirnov (KS), W, and A values in Table 3. So we can say that the proposed distribution is superior among all its competitors as it outperforms all its rivals in fitting COVID-19 data.

8. Additional Points

We propose a novel MKKu distribution generalization, in this paper. We have investigated its statistical properties and determined a linear representation for its PDF that effectively determined the linear representations of the PDF and CDF, moments, moment generating function, and fuzzy reliability function. Numerous classical and Bayesian estimation methods were considered to obtain point estimates for the unknown MKKu parameters , , . The simulation study was done by using the R package, to see and compare the performance of various estimation methods. The MCMC method was used to accomplish this, and we concluded that the Bayesian approach outperforms all the classical approaches considered. We used a COVID-19 data set from the KSA. It was determined that MKKu fitted the data more effectively than most other competing distributions.

9. Future Work

In future, we will work on different kinds of fuzzy reliability function under different kinds of censored data, not a complete sample to study the performance of the estimators under censoring scheme; also we will update the work to find the optimal censoring scheme. We will also go forward to apply accelerated life test on the censored data under different censoring schemes.

Data Availability

All data are available in the paper.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

References

T. A. Anake, P. E. Oguntunde, and O. A. Odetunmibi, “On a fractional beta- distribution,” International Journal of Mathematics and Computation, vol. 26, pp. 26–34, 2015.
View at: Google Scholar
S. Kumar, “Monitoring novel corona virus (COVID-19) infections in India by cluster Analysis,” Annals of Data Science, vol. 7, no. 3, pp. 417–425, 2020.
View at: Publisher Site | Google Scholar
A. Khakharia, V. Shah, S. Jain et al., “Outbreak prediction of covid-19 for dense and populated countries using machine learning,” Annals of Data Science, vol. 8, no. 1, pp. 1–19, 2020.
View at: Publisher Site | Google Scholar
A. Shafiq, S. A. Lone, T. N. Sindhu, Y. El Khatib, Q. M. Al-Mdallal, and T. Muhammad, “A new modified Kies Fréchet distribution: applications of mortality rate of COVID-19,” Results in Physics, vol. 28, Article ID 104638, 2021.
View at: Publisher Site | Google Scholar
T. N. Sindhu, A. Shafiq, and Q. M. Al-Mdallal, “Exponentiated transformation of Gumbel Type-II distribution for modeling COVID-19 data,” Alexandria Engineering Journal, vol. 60, no. 1, pp. 671–689, 2021.
View at: Publisher Site | Google Scholar
Y. X. J. Wáng, “A call for caution in extrapolating chest CT sensitivity for COVID-19 derived from hospital data to patients among general population,” Quantitative Imaging in Medicine and Surgery, vol. 10, no. 3, pp. 798-799, 2020.
View at: Publisher Site | Google Scholar
S. Lalmuanawma, J. Hussain, and L. Chhakchhuak, “Applications of machine learning and artificial intelligence for COVID-19 (SARS-CoV-2) pandemic: a review,” Chaos, Solitons & Fractals, vol. 139, Article ID 110059, 2020.
View at: Publisher Site | Google Scholar
J. Bullock, K. H. Pham, C. S. N. Lam, and M. Luengo-Oroz, “Mapping the landscape of artificial intelligence applications against COVID-19,” 2020, https://arxiv.org/abs/2003.11336.
View at: Google Scholar
A. S. Hassan, M. Elgarhy, and R. Ragab, “Statistical properties and estimation of inverted Topp-Leone distribution,” Journal of Applied Probability and Statistics, vol. 9, no. 2, 2020.
View at: Google Scholar
C. S. Kumar and S. H. S. Dharmaja, “The exponentiated reduced Kies distribution: properties and applications,” Communications in Statistics-Theory and Methods, vol. 46, no. 17, pp. 8778–8790, 2017.
View at: Publisher Site | Google Scholar
S. Dey, M. Nassar, and D. Kumar, “Moments and estimation of reduced Kies distribution based on progressive type-II right censored order statistics,” Hacettepe Journal of Mathematics and Statistics, vol. 48, pp. 332–350, 2019.
View at: Google Scholar
A. A. Al-Babtain, M. K. Shakhatreh, M. Nassar, and A. Z. Afify, “A new modified Kies family: properties, estimation under complete and type-II censored samples, and engineering applications,” Mathematics, vol. 8, no. 8, Article ID 1345, 2020.
View at: Publisher Site | Google Scholar
R. A. R. Bantan, F. Jamal, C. Chesneau, and M. Elgarhy, “Type II Power Topp-Leone generated family of distributions with statistical inference and applications,” Symmetry, vol. 12, no. 1, p. 75, 2020.
View at: Publisher Site | Google Scholar
N. Kunjiratanachot, W. Bodhisuwan, and A. Volodin, “The Topp-Leone generalized exponential power series distribution with applications,” Journal of Probability and Statistics, vol. 16, no. 2, pp. 197–208, 2018.
View at: Google Scholar
P. E. Oguntunde, M. A. Khaleel, M. T. Ahmed, A. O. Adejumo, and O. A. Odetunmibi, “A new generalization of the Lomax distribution with increasing, decreasing, and constant failure rate,” Modelling and Simulation in Engineering, vol. 2017, Article ID 6043169, 12 pages, 2017.
View at: Publisher Site | Google Scholar
G. M. Ibrahim, A. S. Hassan, E. M. Almetwally, and H. M. Almongy, “Parameter estimation of alpha power inverted Topp-Leone distribution with applications,” Intelligent Automation & Soft Computing, vol. 29, no. 2, pp. 353–371, 2021.
View at: Google Scholar
J.-M. Kim, H. Ju, and Y. Jung, “Copula approach for developing a biomarker panel for prediction of dengue hemorrhagic fever,” Annals of Data Science, vol. 7, no. 4, pp. 697–712, 2020.
View at: Publisher Site | Google Scholar
M. Aslam, R. Yousaf, and S. Ali, “Bayesian estimation of transmuted pareto distribution for complete and censored data,” Annals of Data Science, vol. 7, no. 4, pp. 663–695, 2020.
View at: Publisher Site | Google Scholar
E. M. Almetwally, “The odd Weibull inverse Topp–Leone distribution with applications to COVID-19 data,” Annals of Data Science, vol. 9, pp. 1–20, 2021.
View at: Google Scholar
A. S. M. Metwally, A. S. Hassan, E. M. Almetwally, B. M. G. Kibria, and H. M. Almongy, “Reliability analysis of the new exponential inverted topp-leone distribution with applications,” Entropy, vol. 23, no. 12, Article ID 1662, 2021.
View at: Publisher Site | Google Scholar
M. S. Eliwa, M. El-Morshedy, and S. Ali, “Exponentiated odd Chen-G family of distributions: statistical properties, Bayesian and non-Bayesian estimation with applications,” Journal of Applied Statistics, vol. 48, no. 11, pp. 1948–1974, 2021.
View at: Publisher Site | Google Scholar
A. Zellner, “Bayesian and non-Bayesian estimation using balanced loss functions,” Statistical Decision Theory and Related Topics V, Springer, New York, NY, pp. 377–390, 1994.
View at: Publisher Site | Google Scholar
H. Z. Muhammed and E. M. Almetwally, “Bayesian and non-Bayesian estimation for the bivariate inverse Weibull distribution under progressive type-II censoring,” Annals of Data Science, 2020.
View at: Publisher Site | Google Scholar
E. S. A. El-Sherpieny, E. M. Almetwally, and H. Z. Muhammed, “Bayesian and non-Bayesian estimation for the parameter of bivariate generalized Rayleigh distribution based on clayton copula under progressive type-II censoring with random removal,” Sankhya-Series A, pp. 1–38, 2021.
View at: Publisher Site | Google Scholar
R. C. H. Cheng and N. A. K. Amin, “Estimating parameters in continuous univariate distributions with a shifted origin,” Journal of the Royal Statistical Society: Series B, vol. 45, no. 3, pp. 394–403, 1983.
View at: Publisher Site | Google Scholar
R. Kumar Singh, S. Kumar Singh, and U. Singh, “Maximum product spacings method for the estimation of parameters of generalized inverted exponential distribution under Progressive Type II Censoring,” Journal of Statistics & Management Systems, vol. 19, no. 2, pp. 219–245, 2016.
View at: Publisher Site | Google Scholar
S. Basu, S. K. Singh, and U. Singh, “Estimation of inverse lindley distribution using product of spacings function for hybrid censored data,” Methodology and Computing in Applied Probability, vol. 21, no. 4, pp. 1377–1394, 2019.
View at: Publisher Site | Google Scholar
R. Alshenawy, M. Sabry, E. Almetwally, and H. Elomngy, “Product spacing of stress-strength under progressive hybrid censored for exponentiated-gumbel distribution,” Computers, Materials & Continua, vol. 66, no. 3, pp. 2973–2995, 2021.
View at: Publisher Site | Google Scholar
A. Luceño, “Fitting the generalized Pareto distribution to data using maximum goodness-of-fit estimators,” Computational Statistics & Data Analysis, vol. 51, no. 2, pp. 904–917, 2006.
View at: Publisher Site | Google Scholar
S. S. Donighi and S. Khanmohammadi, “A fuzzy reliability model for series-parallel systems,” International Journal of Industrial Engineering, vol. 7, no. 12, pp. 10–18, 2011.
View at: Google Scholar
H.-C. Wu, “Fuzzy reliability estimation using Bayesian approach,” Computers & Industrial Engineering, vol. 46, no. 3, pp. 467–493, 2004.
View at: Publisher Site | Google Scholar
L. A. Zadeh, “Fuzzy sets, fuzzy logic, and fuzzy systems: selected papers by,” Advances in Fuzzy Systems - Applications and Theory, vol. 8, no. 3, pp. 394–432, 1996.
View at: Publisher Site | Google Scholar
G. Chen and T. T. Pham, Introduction to Fuzzy Sets, Fuzzy Logic, and Fuzzy Control Systems, CRC Press, Florida, USA, 2000.
E. S. A. El-Sherpieny and M. A. Ahmed, “On the Kumaraswamy Kumaraswamy distribution,” International Journal of Basic and Applied Sciences, vol. 3, no. 4, pp. 372–381, 2014.
View at: Publisher Site | Google Scholar
F. Jamal, H. M. Reyad, M. A. Nasir, C. Chesneau, M. A. A. Shah, and S. O. Ahmed, “Topp-Leone Weibull-Lomax distribution: properties, regression model and applications,” 2019.
View at: Google Scholar
R. George and S. Thobias, “Marshall-Olkin Kumaraswamy distribution,” International Mathematical Forum, vol. 12, no. 2, pp. 47–69, 2017.
View at: Publisher Site | Google Scholar
https://covid19.who.int/.

Copyright

Copyright © 2022 Fathy H. Riad et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies