A dependent frequency–severity approach to modeling longitudinal insurance claims

In nonlife insurance, frequency and severity are two essential building blocks in the actuarial modeling of insurance claims. In this paper, we propose a dependent modeling framework to jointly examine the two components in a longitudinal context where the quantity of interest is the predictive distribution. The proposed model accommodates the temporal correlation in both the frequency and the severity, as well as the association between the frequency and severity using a novel copula regression. The resulting predictive claims distribution allows to incorporate the claim history on both the frequency and severity into ratemaking and other prediction applications. In this application, we examine the insurance claim frequencies and severities for specific peril types from a government property insurance portfolio, namely lightning and vehicle claims, which tend to be frequent in terms of their count. We discover that the frequencies and severities of these frequent peril types tend to have a high serial correlation over time. Using dependence modeling in a longitudinal setting, we demonstrate how the prediction of these frequent claims can be improved.     

主成分分析法在非寿险保险公司偿付能力影响因素中的应用

本文运用主成分分析法对非寿险保险公司偿付能力诸多影响因素进行分析,简化成几个综合的成分,所得结果为保险监管部门和保险公司制定切实有效的管理策略,提供有价值的参考。     

Selecting appropriate machine learning methods for digital soil mapping

Digital soil mapping (DSM) increasingly makes use of machine learning algorithms to identify relationships between soil properties and multiple covariates that can be detected across landscapes. Selecting the appropriate algorithm for model building is critical for optimizing results in the context of the available data. Over the past decade, many studies have tested different machine learning (ML) approaches on a variety of soil data sets. Here, we review the application of some of the most popular ML algorithms for digital soil mapping. Specifically, we compare the strengths and weaknesses of multiple linear regression (MLR), k-nearest neighbors (KNN), support vector regression (SVR), Cubist, random forest (RF), and artificial neural networks (ANN) for DSM. These algorithms were compared on the basis of five factors: (1) quantity of hyperparameters, (2) sample size, (3) covariate selection, (4) learning time, and (5) interpretability of the resulting model. If training time is a limitation, then algorithms that have fewer model parameters and hyperparameters should be considered, e.g., MLR, KNN, SVR, and Cubist. If the data set is large (thousands of samples) and computation time is not an issue, ANN would likely produce the best results. If the data set is small (<100), then Cubist, KNN, RF, and SVR are likely to perform better than ANN and MLR. The uncertainty in predictions produced by Cubist, KNN, RF, and SVR may not decrease with large datasets. When interpretability of the resulting model is important to the user, Cubist, MLR, and RF are more appropriate algorithms as they do not function as “black boxes.” There is no one correct approach to produce models for predicting the spatial distribution of soil properties. Nonetheless, some algorithms are more appropriate than others considering the nature of the data and purpose of mapping activity.     

On the solution of some (parametric) linear complementarity problems with applications to portfolio selection,structural engineering and actuarial graduation

In this paper we discuss three applications of a class of (parametric) linear complementarity problems arising independently from such diverse areas as portfolio selection, structural engineering and actuarial graduation. After explaining how the complementarity problems emerge in these applications, we perform some analytical comparisons (based on operation counts and storage requirements) of several existing algorithms for solving this class of complementarity problems. We shall also present computational results to support the analytical comparisons. Finally, we deduce some conclusions about the general performance of these algorithms.This research is supported in part by the United States Army under Contract No. DAAG29-75-C-0024, the National Science Foundation under Grant No. MCS75-17385 and Grant ENG77-11136.     

Modeling frequency and severity of claims with the zero-inflated generalized cluster-weighted models

To facilitate applications in general insurance, some extensions are proposed to cluster-weighted models (CWMs). First, we extend CWMs to have generalized cluster-weighted models (GCWMs) by allowing modeling of non-Gaussian distribution of the continuous covariates, as they frequently occur in insurance practice. Secondly, we introduce a zero-inflated extension of GCWM (ZI-GCWM) for modeling insurance claims data with excess zeros coming from heterogeneous sources. Additionally, we give two expectation–optimization (EM) algorithms for parameter estimation given in the proposed models. An appropriate simulation study shows that, for various settings and in contrast to the existing mixture-based approaches, both extended models perform well. Finally, a real data set based on French auto-mobile policies is used to illustrate the application of the proposed extensions.     

Developments in linear and integer programming

In this review we describe recent developments in linear and integer (linear) programming. For over 50 years Operational Research practitioners have made use of linear optimisation models to aid decision making and over this period the size of problems that can be solved has increased dramatically, the time required to solve problems has decreased substantially and the flexibility of modelling and solving systems has increased steadily. Large models are no longer confined to large computers, and the flexibility of optimisation systems embedded in other decision support tools has made on-line decision making using linear programming a reality (and using integer programming a possibility). The review focuses on recent developments in algorithms, software and applications and investigates some connections between linear optimisation and other technologies.     

GAMLSS模型在车险定价中的应用

在系统梳理国内外非寿险产品费率厘定方法的基础上,详细介绍了GAMLSS模型,证明了在位置参数和尺度参数的预测中均引入随机效应的GAMLSS模型可更有效地解释纵向数据中个体间的异质性.最后将GAMLSS模型应用于一组纵向车辆保险数据,计算了先验保费、后验保费、后验风险保费和奖惩因子.实证结果表明,GAMLSS模型不仅可为非寿险产品的定价提供依据,而且使风险分类更加稳定、合理.     

Analytical modelling for static stress analysis of pin-loaded lugs with bush fitting

Pin-loaded connections are widely used in the industry. Connecting rods of reciprocating engines in automotive applications or pin-loaded lugs of landing gears and airframe structures can be cited as examples. These systems are subject to thermomechanical cyclic loadings and fatigue mechanisms are often observed. The calculation of stresses in the lug is crucial in view of a fatigue lifetime analysis. Analytical modelling may be very helpful in the early design stage, by allowing fast parametric studies and the identification of the most influent parameters on the response in stresses. In this paper, we propose analytical models for bush fitting and pin-loading conditions, leading to the complete calculation of the stress distribution in the lug. In each case, the analytical solutions are compared with finite element simulations. In the last part, we perform a sensitivity analysis as an application of the presented analytical tools.     

日本寿险业效率变动实证研究:1998-2008

日本是世界上保险业最发达的国家之一.但自20世纪90年代以来随着泡沫经济的破灭,宏观经济和金融环境的恶化,日本保险业尤其是寿险业进入了停滞与调整期.面对这一情况,日本政府层面和公司层面都推出各种措施,振兴踯躅前行中的日本寿险业.利用数据包络分析(DEA)方法对1998 2008年期间日本全部寿险公司的技术效率、纯技术效率和规模效率及其变动趋势进行了测度,对日本寿险市场的发展演变历程进行全面分析,希望对中国寿险业的发展提供经验借鉴.     

Tensor properties of multilevel Toeplitz and related matrices

A general proposal is presented for fast algorithms for multilevel structured matrices. It is based on investigation of their tensor properties and develops the idea recently introduced by Kamm and Nagy in the block Toeplitz case. We show that tensor properties of multilevel Toeplitz matrices are related to separation of variables in the corresponding symbol, present analytical tools to study the latter, expose truncation algorithms preserving the structure, and report on some numerical results confirming advantages of the proposal.     

Multi-armed bandit-based hyper-heuristics for combinatorial optimization problems

There are significant research opportunities in the integration of Machine Learning (ML) methods and Combinatorial Optimization Problems (COPs). In this work, we focus on metaheuristics to solve COPs that have an important learning component. These algorithms must explore a solution space and learn from the information they obtain in order to find high-quality solutions. Among the metaheuristics, we study Hyper-Heuristics (HHs), algorithms that, given a number of low-level heuristics, iteratively select and apply heuristics to a solution. The HH we consider has a Markov model to produce sequences of low-level heuristics, which we combine with a Multi-Armed Bandit Problem (MAB)-based method to learn its parameters. This work proposes several improvements to the HH metaheuristic that yields a better learning for solving problem instances. Specifically, this is the first work in HHs to present Exponential Weights for Exploration and Exploitation (EXP3) as a learning method, an algorithm that is able to deal with adversarial settings. We also present a case study for the Vehicle Routing Problem with Time Windows (VRPTW), for which we include a list of low-level heuristics that have been proposed in the literature. We show that our algorithms can handle a large and diverse list of heuristics, illustrating that they can be easily configured to solve COPs of different nature. The computational results indicate that our algorithms are competitive methods for the VRPTW (2.16% gap on average with respect to the best known solutions), demonstrating the potential of these algorithms to solve COPs. Finally, we show how algorithms can even detect low-level heuristics that do not contribute to finding better solutions to the problem.     

Linear dynamic approximation theory

We study in detail the behavior of some known learning algorithms. We estimate the sum of the squares of the absolute relative errors of some general linear learning algorithms and the sum of the squares of the coefficients obtained by the perceptron algorithm. We prove the convergence of a statistical learning algorithm. The possibility of applications of this theory to biology is discussed.     

Fast and effective algorithms for the liquid crystal display module (LCM) scheduling problem with sequence-dependent setup time

The liquid crystal display module scheduling problem (LCMSP) is a variation of the classical parallel machines scheduling problem, which has many real-world applications, particular, in the thin film transistor liquid crystal display (TFT-LCD) manufacturing industry. In this paper, we present a case study on the LCMSP, which is taken from a final liquid crystal display module (LCM) shop floor in a TFT-LCD industry. For the case we investigated, the jobs are clustered by their product types and the machine setup time is sequentially dependent on the product types of the jobs processed on the machine. In LCMSP, the objective is to maximize the total profit subject to fulfilling contracted quantities without violating the due date and machine capacity restrictions. The LCMSP can be modelled as a multi-level optimization problem. The sub-problem of LCMSP can be transformed into the vehicle routing problem with time window (VRPTW). One can therefore solve the LCMSP efficiently using existing VRPTW algorithms. We present two new algorithms based on the savings algorithms with some modifications to accommodate the LCMSP. Based on the characteristics of the LCM process, a set of test problems is generated covering most of the real-world applications for test purposes. Computational results and performance comparisons show that the proposed algorithms solved the LCMSP efficiently and near-optimally.     

Combining metaheuristics with mathematical programming, constraint programming and machine learning

During the last years, interest on hybrid metaheuristics has risen considerably in the field of optimization and machine learning. The best results found for many optimization problems in science and industry are obtained by hybrid optimization algorithms. Combinations of optimization tools such as metaheuristics, mathematical programming, constraint programming and machine learning, have provided very efficient optimization algorithms. Four different types of combinations are considered in this paper: (i) Combining metaheuristics with complementary metaheuristics. (ii) Combining metaheuristics with exact methods from mathematical programming approaches which are mostly used in the operations research community. (iii) Combining metaheuristics with constraint programming approaches developed in the artificial intelligence community. (iv) Combining metaheuristics with machine learning and data mining techniques.     

Convergence of the multilevel Full Approximation Scheme including theV-cycle

Summary The multilevel Full Approximation Scheme (FAS ML) is a well-known solver for nonlinear boundary value problems. In this paper we prove local quantitative convergence statements for a class of FAS ML algorithms in a general Hilbertspace setting. This setting clearly exhibits the structure of FAS ML. We prove local convergence of a nested iteration for a rather concrete class of FAS ML algorithms in whichV-cycles and only one Jacobilike pre- and post-smoothing on each level are used.     

加权平衡指数损失函数下的广义信度保费估计

在非寿险中,在索赔经历虽然相互独立,但有时会服从不同的分布.通过考虑保费的目标估计来对风险保费进行了研究,并采用正交投影的方法得到了目标问题的最优解,从而得到了加权平衡指数损失函数下的信度估计.此外,给出了结构参数的无偏估计,并给出了模拟.结果表明,在考虑目标保费的情况下,当选取一个合适的权重,可以得到未来保费的最优估计.     

New approaches to statistical learning theory

We present new tools from probability theory that can be applied to the analysis of learning algorithms. These tools allow to derive new bounds on the generalization performance of learning algorithms and to propose alternative measures of the complexity of the learning task, which in turn can be used to derive new learning algorithms.     

Intelligent systems in finance

Business sectors ranging from banking and insurance to retail, are benefiting from a whole new generation of ‘intelligent’ computing techniques. Successful applications include asset forecasting, credit evaluation, fraud detection, portfolio optimization, customer profiling, risk assessment, economic modelling, sales forecasting and retail outlet location. The techniques include expert systems, rule induction, fuzzy logic, neural networks and genetic algorithms, which in many cases are outperforming traditional statistical approaches. Their key features include the ability to recognize and classify patterns, learning from examples, generalization, logical reasoning from premises, adaptability and the ability to handle data which is incomplete, imprecise and noisy. This paper is the first in a series to appear in Applied Mathematical Finance;here we introduce the reader to the basic concepts of intelligent systems, describe their mode of operation and identify applications of the techniques in real world problem domains. Subsequent papers will concentrate on neural networks, genetic algorithms, fuzzy logic and hybrid systems, and will investigate their history and operation more rigorously.     

A two-dimensional ruin problem on the positive quadrant

In this paper we study the joint ruin problem for two insurance companies that divide between them both claims and premia in some specified proportions (modeling two branches of the same insurance company or an insurance and re-insurance company). Modeling the risk processes of the insurance companies by Cramér-Lundberg processes we obtain the Laplace transform in space of the probability that either of the insurance companies is ruined in finite time. Subsequently, for exponentially distributed claims, we derive an explicit analytical expression for this joint ruin probability by explicitly inverting this Laplace transform. We also provide a characterization of the Laplace transform of the joint ruin time.     

A brief historical introduction to Euler's formula for polyhedra,topology, graph theory and networks

This article is essentially devoted to a brief historical introduction to Euler's formula for polyhedra, topology, theory of graphs and networks with many examples from the real-world. Celebrated Königsberg seven-bridge problem and some of the basic properties of graphs and networks for some understanding of the macroscopic behaviour of real physical systems are included. We also mention some important and modern applications of graph theory or network problems from transportation to telecommunications. Graphs or networks are effectively used as powerful tools in industrial, electrical and civil engineering, communication networks in the planning of business and industry. Graph theory and combinatorics can be used to understand the changes that occur in many large and complex scientific, technical and medical systems. With the advent of fast large computers and the ubiquitous Internet consisting of a very large network of computers, large-scale complex optimization problems can be modelled in terms of graphs or networks and then solved by algorithms available in graph theory. Many large and more complex combinatorial problems dealing with the possible arrangements of situations of various kinds, and computing the number and properties of such arrangements can be formulated in terms of networks. The Knight's tour problem, Hamilton's tour problem, problem of magic squares, the Euler Graeco-Latin squares problem and their modern developments in the twentieth century are also included.