Fitting phase–type scale mixtures to heavy–tailed data and distributions

We consider the fitting of heavy tailed data and distributions with a special attention to distributions with a non–standard shape in the “body” of the distribution. To this end we consider a dense class of heavy tailed distributions introduced in Bladt et al. (Scand. Actuar. J., 573–591 2015), employing an EM algorithm for the maximum likelihood estimation of its parameters. We present methods for fitting to observed data, histograms, censored data, as well as to theoretical distributions. Numerical examples are provided with simulated data and a benchmark reinsurance dataset. Empirical examples show that the methods will in most cases adequately fit both body and tail simultaneously.     

Are plastic yielding and work hardening sensitive to the boundary conditions?

The plane strain shear of a single crystal strip with one active slip system placed in a mixed device with one clamped and one free boundary is considered. Since dislocations pile up against only the clamped boundary, the plastic yielding and work hardening differ essentially from those of a hard device, showing clearly their sensitivity to the boundary conditions. An analytical solution to this problem within continuum dislocation theory is found explicitly which exhibits the energetic and dissipative thresholds for dislocation nucleation, the Bauschinger translational work hardening, and the size effects. (© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)     

Are more suppliers better?: generalising the Guo and Ganeshan procedure

In multiple supplier inventory models, where several suppliers are used to replenish the stock of one item, computation of mean and variance of supplier lead times requires the knowledge of the moments of order statistics from the parent lead time distribution. This article presents a general procedure of finding the moments of supplier lead times in multiple supplier inventory models. The procedure is based on using the Generalised Lambda Distribution (GLD) to approximate the lead time distribution. Numerical examples are provided to validate our procedure. The proposed procedure has the advantage that the computations involved are very simple and can be used for any continuous unimodal lead time distribution.     

What is the role and contribution of models to management and research in the health services? A view from Europe

Modelling in different forms has long been regarded as a cornerstone of health OR as in other fields of application. Models are now tending to become a standard tool in health services management and research. What are the lessons we as operational researchers have learnt during this development? How can health care managers and health service researchers benefit from modelling — with or without OR-analysts to guide them between the pitfalls? After an introductory overview concerning the nature and objectives of modelling, examples will be given of modelling applications from different health service areas in order to illustrate the versatility of the method. Further, the choice of methods and models is discussed with special attention to the problems of interpretation and implementation of results. From this overview some conclusions are drawn with regards to advantages and disadvantages of modelling as tools for policy planning and decision-making in the health area. Finally some observations are made concercing the conditions for the future development in the field.     

The odd log-logistic Topp–Leone G family of distributions: heteroscedastic regression models and applications

We introduce a new class of distributions and provide a comprehensive treatment of its mathematical properties. The maximum likelihood method is discussed to estimate the parameters of the new model by means of Monte-Carlo simulation study. The heteroscedastic regression models with long-term survival are introduced to model data sets with the non homogeneity of the error variances in the presence of cured individuals. The potentiality of the proposed models is illustrated by means of four real data sets.     

View of the world according to Wikipedia: Are we all little Steinbergs?

Saul Steinberg's most famous cartoon “View of the world from 9th Avenue” depicts the world as seen by self-absorbed New Yorkers. By analysing wikipediae of a range of different languages, we find that this particular fish-eye world view is ubiquitous and inherently part of human nature.By measuring the skew in the distribution of locations in different languages we can confirm the validity of plausible quantitative models. These models demonstrate convincingly that people all have similar world views: “We are all little Steinbergs.”Our Steinberg hypothesis allows the world view of specific people to be more accurately modelled; this will allow greater understanding of a person's discourse, either by someone else or automatically by a computer.     

Controllability and synchronizability: Are they related?

In the two last decades the concept of observability has been formally linked to that of embedding in the context of nonlinear dynamics. Such a concept has been shown to play an important role in global modeling, data analysis and filtering, to mention a few examples. Preliminary results suggested that observability, at least in some cases, has some influence in synchronization problems. Could the dual concept of controllability also be important in such problems? In the context of synchronization, in general, the role played by controllability properties may not be as relevant as observability is for data analysis. In this work we compute controllability coefficients analogous to the observability ones, now established in the literature, and evaluate their importance in synchronization problems. Two benchmarks have been used in the simulations: the Rössler and the cord systems. The following schemes were investigated: synchronization to external sinusoidal force, complete replacement, uni- and bi-directional coupling of identical oscillators. The results discussed in this work show that controllability and synchronizability are not related in general.     

On the convolution of Mittag–Leffler distributions and its applications to fractional point processes

We obtain the distribution of the sum of independent Mittag–Leffler (ML) random variables which are not necessarily identically distributed. Firstly we discuss the corresponding known result for independent and identically distributed ML random variables which follows as a special case of our result. Some applications of the obtained result to fractional point processes are also discussed.     

Measurement of Returns to Scale and Damages to Scale for DEA-based operational and environmental assessment: How to manage desirable (good) and undesirable (bad) outputs?

Environmental assessment is increasingly important in preventing various types of pollutions. Data Envelopment Analysis (DEA) has been long used as an operational performance measure, but we have insufficiently explored the use of DEA for environmental assessment. This study explores a new use of DEA for the environmental assessment in which outputs are classified into desirable (good) and undesirable (bad) outputs. Such an output separation is important in the DEA-based environmental assessment. This study extends the use of DEA to the measurement of both Returns to Scale (RTS) for desirable outputs and Damages to Scale (DTS) for undesirable outputs. A Range-Adjusted Measure (RAM) is used as a DEA model for this study because the non-radial model can easily combine the two types of outputs in a unified treatment. All the mathematical features regarding the RAM-based RTS/DTS measurement are first discussed from the operational and environmental performance in a separate treatment. Then, this study combines the two performance measures as a unified measure. The RAM-based RTS/DTS is mathematically explored from the unified measure for operational and environmental performance.     

Regularity criterion to some liquid crystal models and the Landau-Lifshitz equations in R~3

We consider the regularity problem under the critical condition to some liquid crystal models and the Landau-Lifshitz equations. The Serrin type reularity criteria are obtained in the terms of the Besov spaces.     

How good are the proximal point algorithms?

Proximal point algorithms are applicable to a variety of settings in optimization. See Rockafellar, R.T. (1976), and Spingarn, J.E. (1981) for examples. We consider a simple idealized proximal point algorithm using gradient minimization on C² convex functions. This is compared to the direct use of the same gradient method with an appropriate mollifier. The comparison is made by determining estimates of the costrequired to reduce the function to a given precision E. Our object is to assess the potential efficiency of these algorithms even if we do not know how to realize this potential.

We find that for distant starting values, proximal point algorithms are considerably less laborious than a direct method. However there is no essential improvement in the complexity - only in the numerical factors. This negative conclusion holds for the entire family of proximal point algorithms based on the gradient methods of this paper.

The algorithms considered may be important for large scale optimization problems. In applications, the precision e that is desired is usually fixed. Assume this is the case and assume that one is given a family of problems parameterized by the dimension

n. Suppose further that for all n, the condition number Q (defined below) is bounded. Then it will be seen below that for all n sufficiently large our algorithms will require a smaller number of steps than a polynomial algorithm with cost n |Ine|     

Are complex systems hard to evolve?

Evolutionary complexity is measured here by the number of trials/evaluations needed for evolving a logical gate in a nonlinear medium. Behavioral complexity of the gates evolved is characterized in terms of cellular automata behavior. We speculate that hierarchies of behavioral and evolutionary complexities are isomorphic up to some degree, subject to substrate specificity of evolution, and the spectrum of evolution parameters. © 2009 Wiley Periodicals, Inc. Complexity, 2009     

Design Science,the design of systems and Operational Research: back to the future?

The relationship between the science of decision making and the design of systems that support or automate the decision-making process is vital to both academic and practical Operational Research (OR). There is a long-standing need to better understand how OR can be structured so as to contribute to the design of systems, rather than just the science of decisions. This paper articulates the notion of design-oriented OR, particularly within the context of IT-enabled systems, as one approach to addressing this issue. It proposes Design Science (DS) as one way of achieving design-oriented OR, and suggests a reference model for DS as a design-oriented approach to OR. The use and value of DS within the broader remit of management research, and differences with existing systems methodologies, are briefly considered. It concludes by noting that a DS approach would return OR, at least partially, to the values and approach of early OR practitioners, and hence back to the future.     

Factorized time-dependent distributions for certain multiclass queueing networks and an application to?enzymatic processing networks

Motivated by applications in biological systems, we show for certain multiclass queueing networks that time-dependent distributions for the multiclass queue-lengths can have a factorized form which reduces the problem of computing such distributions to a similar problem for related single-class queueing networks. We give an example of the application of this result to an enzymatic processing network.     

Nash bargaining and leader–follower models in water allocation: Application to the Zarrinehrud River basin,Iran

Mathematical models for conflict resolution are very important in integrated water resources and environmental management. This study proposes a new methodology to resolve conflicts among different water users and water suppliers while considering environmental requirements and the system’s constraints. A two-level leader–follower model is applied to maximize the net benefit with the Iran Water Resources Management Company as the leader and agricultural, domestic, and industrial users as followers subject to the system’s constraints. As a comparison, the Nash bargaining solution is also used to find a solution when simultaneous moves are assumed by the participants. The suggested method is then applied to the real case of the Zarrinehrud River basin that is one of the areas facing water shortages in Iran. For the actual optimization, Genetic Algorithm is used in order to avoid local optimum. As the contribution of this study, the results show that benefits for the leader in the leader–follower model increased in comparison with the Nash bargaining solutions.     

Critique of probabilistic models: Application of the Semi‐Markov model to migration

The role of probabilistic models, with special reference to application of the Semi‐Markov model to the analysis of internal migration, is discussed. Advantages of probabilistic models over more conventional methods based on regression analysis are detailed. Probabilistic models have stood apart from the rest of the literature, in migration as well as other areas of application, because they have no substantive content. It is argued that unless their parameters are related to the causal structure and exogenous determinants of the migration process, probabilistic models will be of little practical and scientific use. A strategy for integrating probabilistic models with theoretical and empirical analysis is sketched, to be carried out in the next article.