This document presents a new generalized Lindley distribution (NGLD). The NGLD contains the gamma, exponential, and Lindley distributions as special cases. Statistical properties of the NGLD like the hazard function, moments, and moment generating function are derived. Maximum likelihood estimation is discussed to estimate the parameters of the NGLD. Two real data sets are analyzed to illustrate the usefulness of the new distribution.
In [1] Abdel-Aal has introduced the notions of m-shadow graphs and n-splitting graphs, for all m, n ³ 1.
In this paper, we prove that, the m-shadow graphs for paths and complete bipartite graphs are odd
harmonious graphs for allm³ 1. Also, we prove the n-splitting graphs for paths, stars and symmetric
product between paths and null graphs are odd harmonious graphs for all n³ 1. In addition, we present
some examples to illustrate the proposed theories. Moreover, we show that some families of graphs admit
odd harmonious libeling.
Abstract Quadripartitioned single valued neutrosophic (QSVN) set is a powerful structure where we have four components Truth-T, Falsity-F, Unknown-U and Contradiction-C. And also it generalizes the concept of fuzzy, initutionstic and single valued neutrosophic set. In this paper we have proposed the concept of K-algebras on QSVN, level subset of QSVN and studied some of the results. In addition to this we have also investigated the characteristics of QSVN Ksubalgebras under homomorphism.
Pattern-based classification of demographic sequencesDmitrii Ignatov
We have proposed prefix-based gapless sequential patterns for classification of demographic sequences. In comparison to black-box machine learning techniques, this one provides interpretable patterns suitable for treatment by professional demographers. As for the language, we have used Pattern Structures as an extension of Formal Concept Analysis for the case of complex data like sequences, graphs, intervals, etc.
In [1] Abdel-Aal has introduced the notions of m-shadow graphs and n-splitting graphs, for all m, n ³ 1.
In this paper, we prove that, the m-shadow graphs for paths and complete bipartite graphs are odd
harmonious graphs for allm³ 1. Also, we prove the n-splitting graphs for paths, stars and symmetric
product between paths and null graphs are odd harmonious graphs for all n³ 1. In addition, we present
some examples to illustrate the proposed theories. Moreover, we show that some families of graphs admit
odd harmonious libeling.
Abstract Quadripartitioned single valued neutrosophic (QSVN) set is a powerful structure where we have four components Truth-T, Falsity-F, Unknown-U and Contradiction-C. And also it generalizes the concept of fuzzy, initutionstic and single valued neutrosophic set. In this paper we have proposed the concept of K-algebras on QSVN, level subset of QSVN and studied some of the results. In addition to this we have also investigated the characteristics of QSVN Ksubalgebras under homomorphism.
Pattern-based classification of demographic sequencesDmitrii Ignatov
We have proposed prefix-based gapless sequential patterns for classification of demographic sequences. In comparison to black-box machine learning techniques, this one provides interpretable patterns suitable for treatment by professional demographers. As for the language, we have used Pattern Structures as an extension of Formal Concept Analysis for the case of complex data like sequences, graphs, intervals, etc.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
The idea of metric dimension in graph theory was introduced by P J Slater in [2]. It has been found
applications in optimization, navigation, network theory, image processing, pattern recognition etc.
Several other authors have studied metric dimension of various standard graphs. In this paper we
introduce a real valued function called generalized metric G X × X × X ® R+ d : where X = r(v /W) =
{(d(v,v1),d(v,v2 ),...,d(v,v ) / v V (G))} k Î , denoted d G and is used to study metric dimension of graphs. It
has been proved that metric dimension of any connected finite simple graph remains constant if d G
numbers of pendant edges are added to the non-basis vertices.
Exact Solutions of the Klein-Gordon Equation for the Q-Deformed Morse Potenti...ijrap
In this work, we solve the Klein-Gordon (KG) equation for the general deformed Morse potential with
equal scalar and vector potentials by using the Nikiforov-Uvarov (NU) method, which is based on the
solutions of general second-order linear differential equation with special functions. The energy
eigenvalues and corresponding normalized eigenfunctions are obtained. It is found that the eigenfunctions
can be expressed by the Laguerre polynomials. Our solutions have a good agreement with earlier study.
Polyadic systems and their representations are reviewed and a classification of general polyadic systems is presented. A new multiplace generalization of associativity preserving homomorphisms, a ’heteromorphism’ which connects polyadic systems having unequal arities, is introduced via an explicit formula, together with related definitions for multiplace representations and multiactions. Concrete examples of matrix representations for some ternary groups are then reviewed. Ternary algebras and Hopf algebras are defined, and their properties are studied. At the end some ternary generalizations of quantum groups and the Yang-Baxter equation are presented.
Determination of Optimal Product Mix for Profit Maximization using Linear Pro...IJERA Editor
This paper demonstrates the use of liner programming methods in order to determine the optimal product mix for
profit maximization. There had been several papers written to demonstrate the use of linear programming in
finding the optimal product mix in various organization. This paper is aimed to show the generic approach to be
taken to find the optimal product mix.
In this paper, a Quartic Rank Transmuted Gumbel distribution (QTGD)
to an extended the work of Quartic transmuted distribution families. QTGD increases
the ability of the transmuted distributions to be flexible and facilitate the modelling of
more complex data. We study the main statistical properties of the Quartic transmuted
model, including its hazard rate function, moment-generating function, characteristic
function, quantile function, entropy, and order statistics. Finally, an application of QTGD
,using two real data sets are used to examine the ability to apply it and to observe the
performance of estimation techniques on a Quartic Rank Transmuted Gumbel(QTGD),
Cubic Transmuted Gumbel(CTGD), Transmuted Gumbel(TGD) and Gumbel(GD)
distributions. The observed results showed that QTGD gives better fit than CTGD,TGD,
and GD distributions for the applied data sets.
The Odd Generalized Exponential Log Logistic Distributioninventionjournals
We propose a new lifetime model, called the odd generalized exponential log logistic distribution (OGELLD).We obtain some of its mathematical properties. Some structural properties of the new distribution are studied. The maximum likelihood method is used for estimating the model parameters and the Fisher’s information matrix is derived. We illustrate the usefulness of the proposed model by applications to real lifetime data.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
The idea of metric dimension in graph theory was introduced by P J Slater in [2]. It has been found
applications in optimization, navigation, network theory, image processing, pattern recognition etc.
Several other authors have studied metric dimension of various standard graphs. In this paper we
introduce a real valued function called generalized metric G X × X × X ® R+ d : where X = r(v /W) =
{(d(v,v1),d(v,v2 ),...,d(v,v ) / v V (G))} k Î , denoted d G and is used to study metric dimension of graphs. It
has been proved that metric dimension of any connected finite simple graph remains constant if d G
numbers of pendant edges are added to the non-basis vertices.
Exact Solutions of the Klein-Gordon Equation for the Q-Deformed Morse Potenti...ijrap
In this work, we solve the Klein-Gordon (KG) equation for the general deformed Morse potential with
equal scalar and vector potentials by using the Nikiforov-Uvarov (NU) method, which is based on the
solutions of general second-order linear differential equation with special functions. The energy
eigenvalues and corresponding normalized eigenfunctions are obtained. It is found that the eigenfunctions
can be expressed by the Laguerre polynomials. Our solutions have a good agreement with earlier study.
Polyadic systems and their representations are reviewed and a classification of general polyadic systems is presented. A new multiplace generalization of associativity preserving homomorphisms, a ’heteromorphism’ which connects polyadic systems having unequal arities, is introduced via an explicit formula, together with related definitions for multiplace representations and multiactions. Concrete examples of matrix representations for some ternary groups are then reviewed. Ternary algebras and Hopf algebras are defined, and their properties are studied. At the end some ternary generalizations of quantum groups and the Yang-Baxter equation are presented.
Determination of Optimal Product Mix for Profit Maximization using Linear Pro...IJERA Editor
This paper demonstrates the use of liner programming methods in order to determine the optimal product mix for
profit maximization. There had been several papers written to demonstrate the use of linear programming in
finding the optimal product mix in various organization. This paper is aimed to show the generic approach to be
taken to find the optimal product mix.
In this paper, a Quartic Rank Transmuted Gumbel distribution (QTGD)
to an extended the work of Quartic transmuted distribution families. QTGD increases
the ability of the transmuted distributions to be flexible and facilitate the modelling of
more complex data. We study the main statistical properties of the Quartic transmuted
model, including its hazard rate function, moment-generating function, characteristic
function, quantile function, entropy, and order statistics. Finally, an application of QTGD
,using two real data sets are used to examine the ability to apply it and to observe the
performance of estimation techniques on a Quartic Rank Transmuted Gumbel(QTGD),
Cubic Transmuted Gumbel(CTGD), Transmuted Gumbel(TGD) and Gumbel(GD)
distributions. The observed results showed that QTGD gives better fit than CTGD,TGD,
and GD distributions for the applied data sets.
The Odd Generalized Exponential Log Logistic Distributioninventionjournals
We propose a new lifetime model, called the odd generalized exponential log logistic distribution (OGELLD).We obtain some of its mathematical properties. Some structural properties of the new distribution are studied. The maximum likelihood method is used for estimating the model parameters and the Fisher’s information matrix is derived. We illustrate the usefulness of the proposed model by applications to real lifetime data.
We approach the screening problem - i.e. detecting which inputs of a computer model significantly impact the output - from a formal Bayesian model selection point of view. That is, we place a Gaussian process prior on the computer model and consider the $2^p$ models that result from assuming that each of the subsets of the $p$ inputs affect the response. The goal is to obtain the posterior probabilities of each of these models. In this talk, we focus on the specification of objective priors on the model-specific parameters and on convenient ways to compute the associated marginal likelihoods. These two problems that normally are seen as unrelated, have challenging connections since the priors proposed in the literature are specifically designed to have posterior modes in the boundary of the parameter space, hence precluding the application of approximate integration techniques based on e.g. Laplace approximations. We explore several ways of circumventing this difficulty, comparing different methodologies with synthetic examples taken from the literature.
Authors: Gonzalo Garcia-Donato (Universidad de Castilla-La Mancha) and Rui Paulo (Universidade de Lisboa)
Abstract:
A combination of exponential and Lindley failure rate model is considered and named it as exponential-Lindley
additive failure rate model. In this paper, we studied the distributional properties, central and non-central moments,
estimation of parameters, testing of hypothesis and the power of likelihood ratio criterion about the proposed model.
Key words: Exponential distribution, Lindley distribution, ML estimation, Likelihood ratio type criterion.
Since the advent of the horseshoe priors for regularization, global-local shrinkage methods have proved to be a fertile ground for the development of Bayesian theory and methodology in machine learning. They have achieved remarkable success in computation, and enjoy strong theoretical support. Much of the existing literature has focused on the linear Gaussian case. The purpose of the current talk is to demonstrate that the horseshoe priors are useful more broadly, by reviewing both methodological and computational developments in complex models that are more relevant to machine learning applications. Specifically, we focus on methodological challenges in horseshoe regularization in nonlinear and non-Gaussian models; multivariate models; and deep neural networks. We also outline the recent computational developments in horseshoe shrinkage for complex models along with a list of available software implementations that allows one to venture out beyond the comfort zone of the canonical linear regression problems.
Typically quantifying uncertainty requires many evaluations of a computational model or simulator. If a simulator is computationally expensive and/or high-dimensional, working directly with a simulator often proves intractable. Surrogates of expensive simulators are popular and powerful tools for overcoming these challenges. I will give an overview of surrogate approaches from an applied math perspective and from a statistics perspective with the goal of setting the stage for the "other" community.
We propose a regularized method for multivariate linear regression when the number of predictors may exceed the sample size. This method is designed to strengthen the estimation and the selection of the relevant input features with three ingredients: it takes advantage of the dependency pattern between the responses by estimating the residual covariance; it performs selection on direct links between predictors and responses; and selection is driven by prior structural information. To this end, we build on a recent reformulation of the multivariate linear regression model to a conditional Gaussian graphical model and propose a new regularization scheme accompanied with an efficient optimization procedure. On top of showing very competitive performance on artificial and real data sets, our method demonstrates capabilities for fine interpretation of its parameters, as illustrated in applications to genetics, genomics and spectroscopy.
A COMPARISON OF PARTICLE SWARM OPTIMIZATION AND DIFFERENTIAL EVOLUTIONijsc
Two modern optimization methods including Particle Swarm Optimization and Differential Evolution are
compared on twelve constrained nonlinear test functions. Generally, the results show that Differential
Evolution is better than Particle Swarm Optimization in terms of high-quality solutions, running time and
robustness.
Polyadic systems and their representations are reviewed and a classification of general polyadic systems is presented. A new multiplace generalization of associativity preserving homomorphisms, a 'heteromorphism' which connects polyadic systems having unequal arities, is introduced via an explicit formula, together with related definitions for multiplace representations and multiactions. Concrete examples of matrix representations for some ternary groups are then reviewed.
MSC classes: 16T05, 16T25, 17A42, 20N15, 20F29, 20G05, 20G42, 57T05
A Comparison of Particle Swarm Optimization and Differential Evolutionijsc
Two modern optimization methods including Particle Swarm Optimization and Differential Evolution are compared on twelve constrained nonlinear test functions. Generally, the results show that Differential Evolution is better than Particle Swarm Optimization in terms of high-quality solutions, running time and robustness.
Software Delivery At the Speed of AI: Inflectra Invests In AI-Powered QualityInflectra
In this insightful webinar, Inflectra explores how artificial intelligence (AI) is transforming software development and testing. Discover how AI-powered tools are revolutionizing every stage of the software development lifecycle (SDLC), from design and prototyping to testing, deployment, and monitoring.
Learn about:
• The Future of Testing: How AI is shifting testing towards verification, analysis, and higher-level skills, while reducing repetitive tasks.
• Test Automation: How AI-powered test case generation, optimization, and self-healing tests are making testing more efficient and effective.
• Visual Testing: Explore the emerging capabilities of AI in visual testing and how it's set to revolutionize UI verification.
• Inflectra's AI Solutions: See demonstrations of Inflectra's cutting-edge AI tools like the ChatGPT plugin and Azure Open AI platform, designed to streamline your testing process.
Whether you're a developer, tester, or QA professional, this webinar will give you valuable insights into how AI is shaping the future of software delivery.
JMeter webinar - integration with InfluxDB and GrafanaRTTS
Watch this recorded webinar about real-time monitoring of application performance. See how to integrate Apache JMeter, the open-source leader in performance testing, with InfluxDB, the open-source time-series database, and Grafana, the open-source analytics and visualization application.
In this webinar, we will review the benefits of leveraging InfluxDB and Grafana when executing load tests and demonstrate how these tools are used to visualize performance metrics.
Length: 30 minutes
Session Overview
-------------------------------------------
During this webinar, we will cover the following topics while demonstrating the integrations of JMeter, InfluxDB and Grafana:
- What out-of-the-box solutions are available for real-time monitoring JMeter tests?
- What are the benefits of integrating InfluxDB and Grafana into the load testing stack?
- Which features are provided by Grafana?
- Demonstration of InfluxDB and Grafana using a practice web application
To view the webinar recording, go to:
https://www.rttsweb.com/jmeter-integration-webinar
The Art of the Pitch: WordPress Relationships and SalesLaura Byrne
Clients don’t know what they don’t know. What web solutions are right for them? How does WordPress come into the picture? How do you make sure you understand scope and timeline? What do you do if sometime changes?
All these questions and more will be explored as we talk about matching clients’ needs with what your agency offers without pulling teeth or pulling your hair out. Practical tips, and strategies for successful relationship building that leads to closing the deal.
Neuro-symbolic is not enough, we need neuro-*semantic*Frank van Harmelen
Neuro-symbolic (NeSy) AI is on the rise. However, simply machine learning on just any symbolic structure is not sufficient to really harvest the gains of NeSy. These will only be gained when the symbolic structures have an actual semantics. I give an operational definition of semantics as “predictable inference”.
All of this illustrated with link prediction over knowledge graphs, but the argument is general.
Securing your Kubernetes cluster_ a step-by-step guide to success !KatiaHIMEUR1
Today, after several years of existence, an extremely active community and an ultra-dynamic ecosystem, Kubernetes has established itself as the de facto standard in container orchestration. Thanks to a wide range of managed services, it has never been so easy to set up a ready-to-use Kubernetes cluster.
However, this ease of use means that the subject of security in Kubernetes is often left for later, or even neglected. This exposes companies to significant risks.
In this talk, I'll show you step-by-step how to secure your Kubernetes cluster for greater peace of mind and reliability.
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...UiPathCommunity
💥 Speed, accuracy, and scaling – discover the superpowers of GenAI in action with UiPath Document Understanding and Communications Mining™:
See how to accelerate model training and optimize model performance with active learning
Learn about the latest enhancements to out-of-the-box document processing – with little to no training required
Get an exclusive demo of the new family of UiPath LLMs – GenAI models specialized for processing different types of documents and messages
This is a hands-on session specifically designed for automation developers and AI enthusiasts seeking to enhance their knowledge in leveraging the latest intelligent document processing capabilities offered by UiPath.
Speakers:
👨🏫 Andras Palfi, Senior Product Manager, UiPath
👩🏫 Lenka Dulovicova, Product Program Manager, UiPath
Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024Tobias Schneck
As AI technology is pushing into IT I was wondering myself, as an “infrastructure container kubernetes guy”, how get this fancy AI technology get managed from an infrastructure operational view? Is it possible to apply our lovely cloud native principals as well? What benefit’s both technologies could bring to each other?
Let me take this questions and provide you a short journey through existing deployment models and use cases for AI software. On practical examples, we discuss what cloud/on-premise strategy we may need for applying it to our own infrastructure to get it to work from an enterprise perspective. I want to give an overview about infrastructure requirements and technologies, what could be beneficial or limiting your AI use cases in an enterprise environment. An interactive Demo will give you some insides, what approaches I got already working for real.
Accelerate your Kubernetes clusters with Varnish CachingThijs Feryn
A presentation about the usage and availability of Varnish on Kubernetes. This talk explores the capabilities of Varnish caching and shows how to use the Varnish Helm chart to deploy it to Kubernetes.
This presentation was delivered at K8SUG Singapore. See https://feryn.eu/presentations/accelerate-your-kubernetes-clusters-with-varnish-caching-k8sug-singapore-28-2024 for more details.
Elevating Tactical DDD Patterns Through Object CalisthenicsDorra BARTAGUIZ
After immersing yourself in the blue book and its red counterpart, attending DDD-focused conferences, and applying tactical patterns, you're left with a crucial question: How do I ensure my design is effective? Tactical patterns within Domain-Driven Design (DDD) serve as guiding principles for creating clear and manageable domain models. However, achieving success with these patterns requires additional guidance. Interestingly, we've observed that a set of constraints initially designed for training purposes remarkably aligns with effective pattern implementation, offering a more ‘mechanical’ approach. Let's explore together how Object Calisthenics can elevate the design of your tactical DDD patterns, offering concrete help for those venturing into DDD for the first time!
Elevating Tactical DDD Patterns Through Object Calisthenics
A new generalized lindley distribution
1. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
A new generalized Lindley distribution
Ibrahim Elbatal1 Faton Merovci2* M. Elgarhy3
1.
Institute of Statistical Studies and Research, Departmentof Mathematical Statistics, Cairo University.
2.
3.
Department of Mathematics, University of Prishtina "Hasan Prishtina" Republic of Kosovo
Institute of Statistical Studies and Research, Departmentof Mathematical Statistics, Cairo
1
University.
E-mail: i_elbatal@staff.cu.edu.eg
2
* E-mail of the corresponding author: fmerovci@yahoo.com
3
E-mail: m_elgarhy85@yahoo.com
Abstract
In this paper, we present a new class of distributions called New Generalized Lindley
Distribution(NGLD). This class of distributions contains several distributions such as gamma, exponential
and Lindley as special cases. The hazard function, reverse hazard function, moments and moment
generating function and inequality measures are are obtained. Moreover, we discuss the maximum
likelihood estimation of this distribution. The usefulness of the new model is illustrated by means of two
real data sets. We hope that the new distribution proposed here will serve as an alternative model to other
models available in the literature for modelling positive real data in many areas.
Keywords: Generalized Lindley Distribution; Gamma distribution, Maximum likelihood estimation;
Moment generating function.
1 Introduction and Motivation
In many applied sciences such as medicine, engineering and finance, amongst others, modeling and
analyzing lifetime data are crucial. Several lifetime distributions have been used to model such kinds of data. For
instance, the exponential, Weibull, gamma, Rayleigh distributions and their generalizations ( see, e.g., Gupta and
Kundu, [10]). Each distribution has its own characteristics due specifically to the shape of the failure rate
function which may be only monotonically decreasing or increasing or constant in its behavior, as well as nonmonotone, being bathtub shaped or even unimodal. Here we consider the Lindley distribution which was
introduced by Lindley [13]. Let the life time random variable X has a Lindley distribution with parameter q ,
the probability density function (pdf) of X is given by
q2
g ( x,q ) =
(1 + x)e -q x ; x > 0,q > 0,
q +1
(1)
It can be seen that this distribution is a mixture of exponential (q ) and gamma (2,q ) distributions. The
corresponding cumulative distribution function (cdf) of LD is obtained as
G( x,q ) = 1 -
q + 1 + qx -q x
e , x > 0,q > 0,
q +1
(2)
where q is scale parameter. The Lindley distribution is important for studying stress–strength reliability
modeling. Besides, some researchers have proposed new classes of distributions based on modifications of the
Lindley distribution, including also their properties. The main idea is always directed by embedding former
distributions to more flexible structures. Sankaran [16] introduced the discrete Poisson–Lindley distribution by
combining the Poisson and Lindley distributions. Ghitany et al. [5] have discussed various properties of this
distribution and showed that in many ways Equation (1) provides a better model for some applications than the
exponential distribution. A discrete version of this distribution has been suggested by Deniz and Ojeda [3] having
its applications in count data related to insurance. Ghitany et al. [7, 8] obtained size-biased and zero-truncated
version of Poisson- Lindley distribution and discussed their various properties and applications. Ghitany and AlMutairi [6] discussed as various estimation methods for the discrete Poisson- Lindley distribution. Bakouch et al.
[1] obtained an extended Lindley distribution and discussed its various properties and applications. Mazucheli
and Achcar [14] discussed the applications of Lindley distribution to competing risks lifetime data. Rama and
30
2. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
Mishra [15] studied quasi Lindley distribution. Ghitany et al. [9] developed a two-parameter weighted Lindley
distribution and discussed its applications to survival data. Zakerzadah and Dolati [18] obtained a generalized
Lindley distribution and discussed its various properties and applications.
This paper offers new distribution with three parameter called generalizes the Lindley distribution, this
distribution includes as special cases the ordinary exponential and gamma distributions. The procedure used here
is based on certain mixtures of the gamma distributions. The study examines various properties of the new model.
The rest of the paper is organized as follows: Various statistical properties includes moment, generating function
and inequality measures of the NGL distribution are explored in Section 2. The distribution of the order statistics
is expressed in Section 3. We provide the regression based method of least squares and weighted least squares
estimators in Section 4. Maximum likelihood estimates of the parameters index to the distribution are discussed
in Section 5. Section 6 provides applications to real data sets. Section 7 ends with some conclusions
2 Statistical Properties and Reliability Measures
In this section, we investigate the basic statistical properties, in particular,
generating function and inequality measures for the
rth moment, moment
NGL distribution.
2.1 Density survival and failure rate functions
The new generalized Lindley distribution is denoted as NGLD(a , b ,q ) . This generalized model is
obtained from a mixture of the gamma (a ,q ) and gamma ( b ,q ) distributions as follows:
f ( x,q , a , b ) = pf1 ( x, a ,q ) + (1 - p) f 2 ( x, b ,q )
=
1 éq a +1 x a -1 q b x b -1 ù -q x
+
ê
úe ;a ,q > 0, x > 0. (3)
G( b ) û
1 + q ë G(a )
where
p=
q
1+q
,
f1 ( x, a ,q ) =
q (qx)a -1 -q x
q (qx) b -1 -q x
e and f 2 ( x, b ,q ) =
e .
G(a )
G( b )
The corresponding cumulative distribution function (cdf) of generalized Lindley is given by
F ( x,q , a , b ) =
1 éqg (a ,qx) g ( b ,qx) ù
+
, (4)
1 + q ê G(a )
G( b ) ú
ë
û
where
t
g ( s, t ) = òx s -1e - x dx
0
is called lower incomplete gamma. Also the upper incomplete gamma is given by
¥
G(a , t ) = òxa -1e - x dx
t
For more details about the definition of incomplete gamma, see Wall [20]. Figures 0 and 1 illustrates some of
the possible shapes of the pdf and cdf of the NGL distribution for selected values of the parameters q , a and
b,
respectively.
The survival function associated with (4) is given by
F ( x,q , a , b ) = 1 - F ( x,q , a , b ) = 1 -
1 éqg (a ,qx) g ( b ,qx) ù
, (5)
+
1 + q ê G(a )
G( b ) ú
ë
û
From (??), (4) and (5), the failure (or hazard) rate function (hf) and reverse hazard functions (rhf) of generalized
Lindley distribution are given by
31
3. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
1 éq a +1 x a -1 q b x b -1 ù -q x
+
ê
úe
G( b ) û
f ( x,q , a , b ) 1 + q ë G(a )
, (6)
=
h( x ) =
1 éqg (a , qx) g ( b ,qx) ù
F ( x, q , a , b )
1+
1 + q ê G(a )
G( b ) ú
û
ë
and
Figure 1: The pdf’s of various NGL distributions for values of parameters:
q = 1.5,3,4,5,6,7; a = 0.5,2,3,3.5,4,2.5; with color shapes purple, blue, plum, green, red, black and
darkcyan, respectively.
32
4. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
Figure 2: The cdf’s of various NGL distributions for values of parameters: q = 1,2,3,4,5,6; a = 1,2,3,4,5,6
with color shapes red, green,plum,darcyan, black and purple, respectively.
éq a +1 x a -1 q b x b -1 ù -q x
+
ê
úe
G( b ) û
f ( x, q , a , b )
ë G(a )
t ( x) =
. (7)
=
F ( x, q , a , b )
1 éqg (a ,qx) g ( b ,qx) ù
+
1 + q ê G(a )
G( b ) ú
û
ë
1
1+q
respectively.
Figure 3 illustrates some of the possible shapes of the hazard function of the NGL distribution for
selected values of the parameters q , a and b , respectively.
33
5. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
Figure 3: The hazard’s of various NGL distributions for values of
parameters: q = 1.5,3,4,5,6,7; a = 0.5,2,3,3.5,4,2.5; with color shapes purple, blue, plum, green,
red,blackand darkcyan, respectively.
The following are special cases of the generalized Lindley distribution , GLD(a , b ,q ).
a = 1 and b = 2 , we get the Lindley distribution .
For a = b = l , we get the Gamma distribution with parameter (q , l ).
If a = b = 1 , we get the exponential distribution with parameter q .
1. If
2.
3.
2.2 Moments
Many of the interesting characteristics and features of a distribution can be studied through its moments.
Moments are necessary and important in any statistical analysis, especially in applications. It can be used to
study the most important features and characteristics of a distribution (e.g., tendency, dispersion, skewness and
kurtosis).
Theorem 2.1. If X has GL(f , x) , f = (a ,q , b ) then the rth moment of X is given by the
following
'
m r ( x) =
1
1+q
é G( r + a ) G( r + b ) ù
êq r -1G(a ) + q r G( b ) ú.
û
ë
34
(8)
6. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
Proof:
Let X be a random variable following the GL distribution with parameters q , a and
Expressions for mathematical expectation,
variance and the rth moment on the origin of X can be obtained using the well-known formula
b
.
¥
'
m r ( x) = E ( X r ) = òx r f ( x, f )dx
0
a +1 ¥
q b ¥ r + b -1 -qx ù
1 éq
x r +a -1e -qx dx +
x
e dx ú
ê
1 + q ë G(a ) ò
G( b ) ò
0
0
û
1 é G( r + a ) G( r + b ) ù
.
=
+
1 + q êq r -1G(a ) q r G( b ) ú
ë
û
=
Which completes the proof .
Based on the first four moments of the
(9)
GL distribution, the measures of skewness A(j ) and kurtosis
k (j ) of the GL distribution can obtained as
A(j ) =
m 3 (q ) - 3m1 (q ) m 2 (q ) + 2m13 (q )
[m
2
(q ) - m (q )
2
1
]
3
2
,
and
k (j ) =
m 4 (q ) - 4m1 (q ) m3 (q ) + 6m12 (q ) m2 (q ) - 3m14 (q )
[m (q ) - m
2
1
2
(q )
]
2
.
2.3 Moment generating function
In this subsection we derived the moment generating function of
Theorem (2.2): If X has
following form
M X (t ) =
GL distribution.
GL distribution, then the moment generating function M X (t ) has the
1 é q a +1
qb ù
+
ú.
ê
1 + q ë (q - t )a (q - t ) b û
(10)
Proof.
We start with the well known definition of the moment generating function given by
¥
M X (t ) = E (e ) = òetx f GL ( x, f )dx
tX
0
éq
q b ¥ b -1 -(q -t ) x ù
xa -1e -(q -t ) x dx +
x e
dx ú
ê
G(a ) ò
G( b ) ò
0
0
ë
û
a +1
b
ù
1 é q
q
=
+
ê
a
b ú
1 + q ë (q - t )
(q - t ) û
=
1
1+q
a +1 ¥
Which completes the proof.
In the same way, the characteristic function of the
i = - 1 is the unit imaginary number.
35
(11)
GL distribution becomes j (t ) = M X (it ) where
X
7. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
2.4 Inequality Measures
Lorenz and Bonferroni curves are the most widely used inequality measures in income and wealth
distribution (Kleiber, 2004). Zenga curve was presented by Zenga [19]. In this section, we will derive Lorenz,
Bonferroni and Zenga curves for the GL distribution. The Lorenz, Bonferroni and Zenga curves are defined by
t
é g (a + 1,qt ) g ( b + 1,qt ) ù
ê G(a ) + qG( b ) ú
û.
=ë
bù
E( X )
é
êa + q ú
ë
û
òxf ( x)dx
LF ( x) =
0
(12)
t
òxf ( x)dx
BF ( x) =
0
E ( X ) F ( x)
=
LF ( x)
F ( x)
é g (a + 1,qt ) g ( b + 1,qt ) ù
(1 + q ) ê
+
qG( b ) ú
ë G(a )
û,
=
b ù éqg (a ,qx) g ( b ,qx) ù
é
êa + q ú ê G(a ) + G( b ) ú
ûë
ë
û
(13)
m - ( x)
,
m + ( x)
(14)
and
AF ( x) = 1 where
t
m - ( x) =
é g (a + 1,qt ) g ( b + 1,qt ) ù
ê G(a ) + qG( b ) ú
û
=ë
bù
E( X )
é
êa + q ú
û
ë
òxf ( x)dx
0
and
¥
+
m ( x) =
òxf ( x)dx
t
1 - F ( x)
1 é G(a + 1, qt ) G( b + 1, qt ) ù
+
qG( b ) ú
(1 + q ) ê G(a )
ë
û.
=
1 éqg (a ,qx) g ( b ,qx) ù
1+
1 + q ê G(a )
G( b ) ú
ë
û
respectively.
The mean residual life (mrl) function computes the expected remaining survival time of a subject given
survival up to time x . We have already defined the mrl as the expectation of the remaining survival
time given survival up to time x(see Frank Guess and Frank Proschan [4].
3 Distribution of the order statistics
rth order statistic of the GL
distribution, also, the measures of skewness and kurtosis of the distribution of the rth order statistic in a sample
In this section, we derive closed form expressions for the pdfs of the
36
8. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
n for different choices of n; r are presented in this section. Let X 1 , X 2 ,..., X n be a simple random
sample from GL distribution with pdf and cdf given by (??) and (4), respectively.
Let X 1 , X 2 ,..., X n denote the order statistics obtained from this sample. We now give the probability
density function of X r:n , say f r:n ( x, f ) and the moments of X r:n , r = 1,2,..., n . Therefore, the measures of
skewness and kurtosis of the distribution of the X r:n are presented. The probability density function of X r:n is
of size
given by
1
[F ( x, f ))]r -1[1 - F ( x,f ))]n-r f ( x,f )) (15)
B(r , n - r + 1)
where f ( x, f )) and F ( x, f )) are the pdf and cdf of the GL distribution given by (3) and (4), respectively,
and B (.,.) is the beta function, since 0 < F ( x, f )) < 1 , for x > 0 , by using the binomial series expansion of
f r:n ( x, F) =
[1 - F ( x,f ))]n-r , given by
j
n-rö
÷[F ( x, f ))] ,
÷
è j ø
ç
[1 - F ( x, f ))]n-r = å(-1) j æ
ç
n-r
j =0
(16)
we have
n-r
æn - rö
r + j -1
(17)
f ( x, f )),
f r:n ( x, f )) = å(-1) j ç
ç j ÷[F ( x, F)]
÷
j =0
è
ø
substituting from (??) and (4) into (17), we can express the k th ordinary moment of the rth order statistics X r:n
k
say E ( X r:n ) as a liner combination of the
kth moments of the GL distribution with different shape parameters.
Therefore, the measures of skewness and kurtosis of the distribution of
X r:n can be calculated.
4 Least Squares and Weighted Least Squares Estimators
GL
In this section we provide the regression based method estimators of the unknown parameters of the
distribution, which was originally suggested by Swain, Venkatraman and Wilson [17] to estimate the
parameters of beta distributions. It can be used some other cases also. Suppose
size
Y1 ,..., Yn is a random sample of
n from a distribution function G(.) and suppose Y(i ) ; i = 1,2,..., n denotes the ordered sample. The
proposed method uses the distribution of G (Y(i ) ) . For a sample of size
n , we have
j
j (n - j + 1)
,V (G(Y( j ) ) ) =
n +1
(n + 1) 2 (n + 2)
j (n - k + 1)
; for j < k ,
andCov(G(Y( j ) ), G(Y( k ) ) ) =
(n + 1) 2 (n + 2)
E (G(Y( j ) ) ) =
see Johnson, Kotz and Balakrishnan [11]. Using the expectations and the variances, two variants of the least
squares methods can be used.
Method 1 (Least Squares Estimators) . Obtain the estimators by minimizing
2
j ö
æ
åç G(Y( j ) - n + 1 ÷ ,
ø
j =1 è
with respect to the unknown parameters. Therefore in case of GL distribution the least squares estimators of
ˆ
ˆ
ˆ
a ,q , and b , say a LSE , q LSE and b LSE respectively, can be obtained by minimizing
n
é 1 éqg (a ,qx ) g ( b ,qx ) ù
j ù
åê1 + q ê G(a )( j ) + G(b )( j ) ú - n + 1ú
j =1 ë
û
ë
û
n
37
2
9. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
with respect to
www.iiste.org
a , q , and b .
Method 2 (Weighted Least Squares Estimators). The weighted least squares estimators can be obtained by
minimizing
2
j ö
æ
åw j ç G(Y( j ) - n + 1 ÷ ,
è
ø
j =1
n
with respect to the unknown parameters, where
wj =
Therefore, in case of
· ·
a WLSE ,q WLSE ,
1
(n + 1) 2 (n + 2)
=
.
V (G(Y( j ) ) )
j (n - j + 1)
GL distribution the weighted least squares estimators of a , q , and b , say
·
and
b WLSE
respectively, can be obtained by minimizing
é 1
åw j ê1 + q
j =1
ë
n
éqg (a ,qx( j ) ) g ( b ,qx( j ) ) ù
j ù
+
ú
úê
G( b ) û n + 1û
ë G(a )
2
with respect to the unknown parameters only.
5 Maximum Likelihood Estimators
GL distribution. Let
F = (a ,q , b ) , in order to estimate the parameters a ,q , and b of GL distribution, let x1 ,..., xn be a
random sample of size n from GL(a ,q , b , x) then the likelihood function can be written as
a +1 a -1
b b -1
n
1 éq x(i ) q x(i ) ù -qx(i )
+
L(a ,q , b , x ) = Õ
ê
úe
(i )
G( b ) ú
i =1 1 + q ê G(a )
ë
û
In this section we consider the maximum likelihood estimators (MLE’s) of
T
n
n
æ 1 ö
=ç
÷ e
è1+q ø
n
-q
å
x( i )
i =1
(G(a )G(b ))-n
(
´ Õ G( b )q a +1 x(ai )-1 + G(a )q b x(bi )-1
)
(18)
i =1
By accumulation taking logarithm of equation (18) , and the log- likelihood function can be written as
n
log L = -n log (1 + q ) - q åxi - n log G(a ) - n log G( b )
i =1
n
(
+ å log G( b )q a +1 x(ai )-1 + G(a )q b x(bi )-1
)
(19)
i =1
Differentiating log L with respect to each parameter
a ,q ,
and
obtain maximum likelihood estimates. The partial derivatives of
score function is given by
b and setting the result equals to zero, we
log L with respect to each parameter or the
æ ¶ log L ¶ log L ¶ log L ö
U n (F ) = ç
ç ¶q , ¶a , , ¶b ÷
÷
è
ø
where
n
¶ log L
-n
=
- åxi
(1 + q ) i =1
¶q
38
10. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
n
+å
www.iiste.org
(a + 1)G( b )q a x(ai )-1 + bG(a )q b -1 x(bi )-1
(G(b )q
i =1
x(i ) + G(a )q b x(bi )-1 )
a +1 a -1
= 0,
'
n q a +1 xa -1 log (qx ) + G (a )q b x b -1
¶ log L
(i )
(i )
i
= -ny (a ) + å
= 0,
a +1 a -1
b b -1
¶a
x(i ) + G(a )q x(i ) )
i =1 (G( b )q
(20)
and
'
n q a +1 xa -1G ( b ) + q b x b -1 log (qx )
¶ log L
i
(i )
(i )
= -ny ( b )å
a +1 a -1
b b -1
x(i ) + G(a )q x(i ) )
¶b
i =1 (G( b )q
= 0.
(21)
where y (.) is the digamma function. By solving this nonlinear system of equations (20) - (21), these solutions
·
will yield the ML estimators for
a
·
·
q
,
b
and
. For the three parameters generalized Lindley
distribution GL(a ,q , b , x) pdf all the second order derivatives exist. Thus we have the inverse dispersion
matrix is given by
ˆ
éæ a ö æVaa
ˆ
æa ö
ç ÷
êç ÷ ç ˆ
ˆ
ç q ÷ : N êç q ÷, ç Vqa
çb ÷
êç b ÷ çV
ˆ
çˆ
è ø
ëè ø è ba
V
-1
ˆ
Vab öù
÷
ˆ ú
Vqb ÷ú.
ˆ ÷
Vbb ÷ú
øû
ˆ
Vaq
ˆ
Vqq
ˆ
Vbq
(22)
éVaa Vaq Vab ù
ê
ú
= - E êVqa Vqq Vqb ú
êVba Vbq Vbb ú
ë
û
(23)
where
Vaa =
¶2L
¶2L
¶2L
,Vqq = 2 ,Vbb =
,
¶a 2
¶q
¶b 2
Vaq =
¶2L
¶2L
¶2L
,Vba =
,Vbq =
¶a¶q
¶a¶b
¶b¶q
n
Vaa = -ny ' (a ) + å( Ai + Bi )
i =1
G(b )(ln (q )) q a +1 xi
2
Ai =
a -1
+ 2G(b )q a +1 ln (q )xi ln (xi ) + G(b )q a +1 xi
a -1
b -1
G(b )q a +1 xi + G(a )q b xi
a -1
y ' (a )G(a )q b xi b -1 + (Y (a ))2 G(a )q b xi b -1
+
a -1
b -1
G(b )q a +1 xi + G(a )q b xi
Bi
(G(b )q
=
a +1
ln (q )xi
a -1
+ G(b )q a +1 xi
(G(b )q
xi
a -1
ln (xi ) + Y (a )G(a )q b xi
+ G(a )q b xi
(Ci + Di )
2
i =1 (G(b )
q a +1 xia -1 + G(a )q b xi b -1 ) q
n
Vaq = å
G(b )q a +1 xi
a +1
a -1
a -1
39
)
b -1 2
)
b -1 2
a -1
(ln (xi ))2
11. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
Ci = ln (q )a G(a )q b xi
Di = G(a )q b xi
b -1
n
i
n
-å
+ G(a )q b xi
b -1
+ ln (xi )a G(a )q b xi
+y (a )G(a )q b b xi
Vab = å
b -1
b -1
www.iiste.org
ln (q ) + G(b )q a +1 xi
+ ln (xi )G(a )q b xi
b -1
- ln (q )G(a )q b b xi
b -1
a -1
b -1
- ln (xi )G(a )q b b xi
b -1
- Y(a )G(a )q b xi a - Y(a )G(a )q b xi
b -1
y (b )G(b )q a +1 ln (q xi )xia -1 +y (a )G(a )q b xi b -1 ln (q xi )
a -1
b -1
G(b )q a +1 xi + G(a )q b xi
(G(b )q
a +1
xi
a -1
ln (q xi ) + Y(a )G(a )q b xi
(G(b )q
i
a +1
xi
b -1
a -1
)(Y(b )G(b )q
+ G(a )q xi
b
a +1
)
xi
a -1
+ G(a )q b xi
b -1
)
ln (q xi )
b -1 2
y (b )G(b )q a +1 (a + 1)xia -1 + G(a )q b b xi b -1 ln (q xi ) + G(a )q b xi b -1
a -1
b -1
G(b )q a +1 xi + G(a )q b xi q
i =1
n
Vqb = å
(q
-å
n
(
G(b )(a + 1)xi
a +1
a -1
)
+ G(a )q b b xi
(G(b )q
i =1
a +1
b -1
xi
)(y (b )G(b )q
a -1
+ G(a )q b xi
a +1
xi
a -1
)q
+ G(a )q b xi
b -1
n
G(b )q a +1 xi a 2 + G(b )q a +1 xi a + G(a )q b b 2 xi - G(a )q b b xi
n
Vqq =
+
(1 + q )2 å
q 2 G(b )q a +1 xia -1 + G(a )q b xi b -1
i =1
a -1
(q
-å
n
i =1
a -1
(
G(b )(a + 1)xi
a +1
(
q G(b )q
2
a +1
xi
a -1
a -1
+ G(a )q b b xi
+ G(a )q xi
b
)
ln (q xi )
b -1 2
b -1
b -1
)
)
b -1 2
)
b -1 2
y (b )G(b )q xi + (y (b ))2 G(b )q a +1 xia -1
Vbb = -ny ( b ) + å
a -1
b -1
G(b )q a +1 xi + G(a )q b xi
i =1
n
a -1
a +1
'
'
n
+å
G(a )q b (ln (q )) xi
2
b -1
+ 2G(a )q b xi ln (xi ) ln (q ) + (ln (xi )) G(a )q b xi
a -1
b -1
G(b )q a +1 xi + G(a )q b xi
a -1
+ G(a )q b xi
b -1
i =1
(Y(b )G(b )q
-å
(G(b )q
n
a +1
i =1
xi
a +1
xi
a -1
b -1
+ G(a )q b xi
2
)
ln (q xi )
)
b -1
2
b -1 2
By solving this inverse dispersion matrix these solutions will yield asymptotic variance and covariances of these
ML estimators for
q,
a ,q
and
b . Using (22), we approximate 100(1 - g )%
confidence intervals for
a , b , and
are determined respectively as
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
a ± z g Vaa ,q ± z g Vqq and b ± z g Vbb
2
where zg is the upper
2
2
100g the percentile of the standard normal distribution.
Using R we can easily compute the Hessian matrix and its inverse and hence the standard errors and
asymptotic confidence intervals.
We can compute the maximized unrestricted and restricted log-likelihood functions to construct the
likelihood ratio (LR) test statistic for testing on some the new generalized Lindley sub-models. For example, we
can use the LR test statistic to check whether the new generalized Lindley distribution for a given data set is
statistically superior to the Lindley distribution. In any case, hypothesis tests of the type H 0 : j = j0 versus
H 0 : j ¹ j0 can be performed using a LR test. In this case, the LR test statistic for testing H 0 versus H1 is
40
b -1
12. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
ˆ
ˆ
ˆ
ˆ
w = 2(l(j; x) - l(j0 ; x)) , where j and j 0 are the MLEs under H1 and H 0 , respectively. The statistic w
2
is asymptotically (as n ® ¥ ) distributed as c k , where k is the length of the parameter vector j of interest.
The LR test rejects
H 0 if w > c k2;g , where c k2;g denotes the upper 100g % quantile of the c k2 distribution.
6 Applications
In this section, we use two real data sets to show that the beta Lindley distribution can be a better model
than one based on the Lindley distribution.
Data set 1: The data set given in Table 1 represents an uncensored data set corresponding to remission
times (in months) of a random sample of 128 bladder cancer patients reported in Lee and Wang [12]:
Table 1: The remission times (in months) of bladder cancer patients
0.08
0.52
2.09
4.98
3.48
6.97
4.87
9.02
6.94
13.29
8.66
0.40
13.11
2.26
23.63
3.57
0.20
5.06
2.23
7.09
0.82
0.62
0.39
0.96
0.19
0.66
0.40
0.26
0.31
0.73
0.51
3.82
10.34
36.66
2.75
11.25
3.02
11.98
4.51
2.07
2.54
5.32
14.83
1.05
4.26
17.14
4.34
19.13
6.54
3.36
3.70
7.32
34.26
2.69
5.41
79.05
5.71
1.76
8.53
6.93
5.17
10.06
0.90
4.23
7.63
1.35
7.93
3.25
12.03
8.65
7.28
14.77
2.69
5.41
17.12
2.87
11.79
4.50
20.28
12.63
9.74
32.15
4.18
7.62
46.12
5.62
18.10
6.25
2.02
22.69
14.76
2.64
5.34
10.75
1.26
7.87
1.46
8.37
3.36
5.49
26.31
3.88
7.59
16.62
2.83
11.64
4.40
12.02
6.76
0.81
5.32
10.66
43.01
4.33
17.36
5.85
2.02
12.07
Data set 2: The following data represent the survival times (in days) of 72 guinea pigs infected with
virulent tubercle bacilli, observed and reported by Bjerkedal [2]. The data are as follows:
0.1, 0.33, 0.44, 0.56, 0.59, 0.72, 0.74, 0.77, 0.92, 0.93, 0.96, 1, 1, 1.02, 1.05, 1.07, 1.07, 1.08, 1.08, 1.08,
1.09, 1.12, 1.13, 1.15, 1.16, 1.2, 1.21, 1.22, 1.22, 1.24, 1.3, 1.34, 1.36, 1.39, 1.44, 1.46, 1.53, 1.59, 1 .6, 1.63, 1.63,
1.68, 1.71, 1.72, 1.76, 1.83, 1.95, 1.96, 1.97, 2.02, 2.13, 2.15, 2.16, 2.22, 2.3, 2.31, 2.4, 2.45, 2.51, 2.53, 2.54,
2.54, 2.78, 2.93, 3.27, 3.42, 3.47, 3.61, 4.02, 4.32, 4.58, 5.55
Table 2: The ML estimates, standard error and Log-likelihood for data set 1
Model
Lindley
NGLD
ML Estimates
Standard Error
0.012
ˆ
a = 4.679
ˆ = 1.324
b
ˆ
The variance covariance matrix I (l )
for data set 1 is computed as
-1
-LL
419.529
0.035
ˆ
q = 0.196
ˆ
q = 0.18
412.750
1.308
0.171
of the MLEs under the new generalized Lindley distribution
æ 0.001 0.031 0.005 ö
ç
÷
.ç 0.031 1.711 0.140 ÷
ç 0.005 0.140 0.029 ÷
è
ø
41
13. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
Thus,
the
variances
of
the
MLE
www.iiste.org
of
q ,a
and
b
is
ˆ
ˆ
var(q ) = 0.001, var(a ) = 1.711 and
ˆ
95% confidence intervals
var(b ) = 0.0295. Therefore,
[0.113,0.252], [2.115,7.243] and [0.987,1.661] respectively.
for
q ,a
and
b
are
In order to compare the two distribution models, we consider criteria like - 2l , AIC (Akaike
information criterion), AICC (corrected Akaike information criterion), BIC(Bayesian information criterion) and
K-S(Kolmogorov-Smirnov test) for the data set. The better distribution corresponds to smaller - 2l , AIC and
AICC values:‘
AIC = 2k - 2l,
AICC = AIC +
2k (k + 1)
,
n - k -1
BIC = k * log (n) - 2l and K - S = sup | Fn ( x) - F ( x) |
x
1 n
where Fn ( x) = åI x £ x is empirical distribution function, F (x) is comulative distribution function, k is
n i =1 i
the number of parameters in the statistical model, n the sample size and l is the maximized value of the loglikelihood function under the considered model.
Table 3: The AIC, AICC, BIC and K-S of the models based on data set 1
Model
Lindley
NGLD
-2LL
839.04
825.501
AIC
841.06
831.501
The LR test statistic to test the hypotheses
is
w = 13.539 > 5.991 = c
2
2;0.05,
AICC
841.091
831.694
BIC
843.892
840.057
K-S
0.074
0.116
/
/
H 0 : a = b = 1 versus H1 : a = 1 Ú b = 1 for data set 1
so we reject the null hypothesis.
Figure 4: Estimated densities of the models for data set 1.
42
14. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
Figure 5: Estimated cumulative densities of the models for data set 1.
Figure 6: P-P plots for fitted Lindley and the NGLD for data set 1.
43
15. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
Table 4: The ML estimates, standard error and Log-likelihood for data set 2
Model
Lindley
ML Estimates
St. Error
0.076
ˆ
q = 0.868
ˆ
q = 1.861
NGLD
94.182
0.489
1.238
ˆ
a = 3.585
ˆ = 2.737
b
ˆ
The variance covariance matrix I (l )
computed as
-1
-l
106.928
0.554
of the MLEs under the beta Lindley distribution for data set is
0.569 - 0.001 ö
æ 0.239
÷
ç
1.532 - 0.154 ÷
ç 0.569
ç - 0.001 - 0.154 0.307 ÷
ø
è
ˆ
ˆ
Thus, the variances of the MLE of q , a and b is var(q ) = 0.239,var(a ) = 0.239 and
ˆ
95% confidence
Therefore,
intervals
for
and
are
q ,a
var(b ) = 0.307.
b
[0.901,2.819], [1.158,6.011] and [1.651,3.823] respectively.
Table 5: The AIC, AICC, BIC and K-S of the models based on data set 2
Model
Lindley.
NGLD
- 2l
213.857
188.364
AIC
215.857
194.364
The LR test statistic to test the hypotheses
2
w = 25.493 > 5.991 = c 2;0.05,
AICC
215.942
194.722
BIC
218.133
201.194
K-S
0.232
0.075
/
/
H 0 : a = b = 1 versus H1 : a = 1 Ú b = 1 for data set 2 is
so we reject the null hypothesis. Tables 2 and 4 shows parameter MLEs to
each one of the two fitted distributions for data set 1 and 2, Tables 3 and 5 shows the values of - 2 log( L), AIC,
AICC, BIC and K-S values. The values in Tables 3 and 5, indicate that the new generalized Lindley distribution
is a strong competitor to other distribution used here for fitting data set 1 and data set 2. A density plot compares
the fitted densities of the models with the empirical histogram of the observed data (Fig. 3 and 5). The fitted
density for the new generalized Lindley model is closer to the empirical histogram than the fits of the Lindley
model.
44
16. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
Figure 7: Estimated densities of the models for data set 2.
Figure 8: Estimated cumulative densities of the models for data set 2.
45
17. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
Figure 9: P-P plots for fitted Lindley and the NGLD for data set 2.
7 Conclusion
Here, we propose a new model, the so-called the new generalized Lindley distribution which extends
the Lindley distribution in the analysis of data with real support. An obvious reason for generalizing a standard
distribution is because the generalized form provides larger flexibility in modelling real data. We derive
expansions for the moments and for the moment generating function. The estimation of parameters is
approached by the method of maximum likelihood, also the information matrix is derived. We consider the
likelihood ratio statistic to compare the model with its baseline model. Two applications of the new generalized
Lindley distribution to real data show that the new distribution can be used quite effectively to provide better fits
than the Lindley distribution.
References
Bakouch HS, Al-Zahrani BM, Al-Shomrani AA, Marchi VAA, Louzada F (2012). An extended Lindley
distribution. J. Korean Stat. Soc. 41(1), 75-85.
Bjerkedal, T. (1960). Acquisition of Resistance in Guinea Pies infected with Different Doses of Virulent Tubercle
Bacilli. American Journal of Hygiene, 72(1), 130-48.
Deniz EG, Ojeda EC (2011). The discrete Lindley distribution Properties and Applications. J. Stat. Comput.
Simul.81(11), 1405- 1416.
Frank Guess and Frank Proschan. Mean residual life: theory and applications. (1985). Technical Report 85-178,
North Carolina State University and Florida State University, Tallahassee,Florida, June 1985. Ghitany ME,
Atieh B, and Nadarajah S (2008a). Lindley distribution and its Applications. Math. Comput. Simul. 78(4):493506.
Ghitany ME, and Al-Mutairi DK (2009). Estimation methods for the discrete Poisson- Lindley distribution. J.
Stat. Comput. Simul. 79(1):1-9.
Ghitany ME, and Al-Mutairi DK (2008b). Size-biased Poisson-Lindley distribution and its Applications. MetronInt. J. Stat. LXVI(3): 299 - 311.
Ghitany ME, Al-Mutairi DK, and Nadarajah S (2008c). Zero-truncated Poisson-Lindley distribution and its
Applications. Math. Comput. Simul. 79(3):279-287.
Ghitany ME, Al-qallaf F, Al-Mutairi DK, and Hussain H A (2011). A two parameter weighted Lindley
distribution and its applications to survival data. Math. Comput. Simul. 81(6), 1190-1201.
46
18. Mathematical Theory and Modeling
ISSN 2224-5804 (Paper) ISSN 2225-0522 (Online)
Vol.3, No.13, 2013
www.iiste.org
Gupta, R.D., and Kundu. D (1999). Generalized exponential distributions. Australian and New Zealand Journal
of Statistics. 41, 173 –188.
Johnson, N., Kotz, S. and Balakrishnan,N. (1995). Continuous Univariate Distribution, V(2), 2nd edition, New
York, Wiley.
Lee E.T. and Wang J.W.,(2003).Statistical Methods for Survival Data Analysis, 3rd ed.,Wiley, NewYork.
Lindley DV (1958). Fiducial distributions and Bayes’ theorem. J. Royal Stat. Soc. Series B, 20:102-107.
Mazucheli J, and Achcar J. A (2011). The Lindley distribution applied to competing risks lifetime data.
Compmajtut. Methods Programs Biomed. 104(2):188-192.
Rama S and Mishra A (2013). A quasi Lindley distribution. African Journal of Mathematics and Computer
Science Research. 6(4), 64-71.
Sankaran M (1970). The discrete Poisson-Lindley distribution. Biometrics pp. 145-149.
Swain, J , Venkatraman,S and Wilson, J (1988) Least squares estimation of distribution function in Johnson’s
translation system. Journal of Statistical Computation and Simulation, 29 , 271 - 297.
Zakerzadah H, Dolati A (2010). Generalized Lindley distribution. J. Math. Ext. 3(2):13-25.
Zenga M.,(2007). Inequality curve and inequality index based on the ratios between lower and upper arithmetic
means. Statistica e Applicazioni 4: 3–27
Wall, H. S., (1948). Analytic Theory of Continued Fractions. New York: Chelsea.
47
19. This academic article was published by The International Institute for Science,
Technology and Education (IISTE). The IISTE is a pioneer in the Open Access
Publishing service based in the U.S. and Europe. The aim of the institute is
Accelerating Global Knowledge Sharing.
More information about the publisher can be found in the IISTE’s homepage:
http://www.iiste.org
CALL FOR JOURNAL PAPERS
The IISTE is currently hosting more than 30 peer-reviewed academic journals and
collaborating with academic institutions around the world. There’s no deadline for
submission. Prospective authors of IISTE journals can find the submission
instruction on the following page: http://www.iiste.org/journals/
The IISTE
editorial team promises to the review and publish all the qualified submissions in a
fast manner. All the journals articles are available online to the readers all over the
world without financial, legal, or technical barriers other than those inseparable from
gaining access to the internet itself. Printed version of the journals is also available
upon request of readers and authors.
MORE RESOURCES
Book publication information: http://www.iiste.org/book/
Recent conferences: http://www.iiste.org/conference/
IISTE Knowledge Sharing Partners
EBSCO, Index Copernicus, Ulrich's Periodicals Directory, JournalTOCS, PKP Open
Archives Harvester, Bielefeld Academic Search Engine, Elektronische
Zeitschriftenbibliothek EZB, Open J-Gate, OCLC WorldCat, Universe Digtial
Library , NewJour, Google Scholar