Multi–criteria Classification Methods in Financial and Banking Decisions

Over the past two decades, the quantitative analysis of financial and banking decisions has gained significant interest among researchers and practitioners. A significant part of the research conducted in this field focused on the development of analytical models that can be used in evaluating the alternative ways of action in financial and banking problems. Typically, this evaluation involves the choice of the best alternative, the ranking of the alternatives from the best to the worst ones, or their classification into predefined homogenous classes. This paper is focused on the classification approach illustrating the use of multi–criteria decision aid (MCDA) classification methods in making financial and banking decisions. Three MCDA approaches (the UTADIS method, the ELECTRE TRI method, and the rough set approach) are applied in financial and banking problems, such as business failure prediction, credit–risk assessment, and portfolio selection and management. A comparison is also performed with linear and quadratic discriminant analysis, and logit analysis.     

博弈论框架下的二阶多智能体系统领导者选择算法

随着人工智能的发展,多智能体系统中智能体的异质性、工作环境的复杂性、系统目标的多样性,给多智能体系统动态性能的分析带来了挑战.同时,也催生了新的控制策略和优化方法.博弈论作为一种研究社会系统中智能体决策过程的经典工具,如今已被应用到了多智能体系统研究领域.本文主要针对二阶多智能体系统编队过程中领导者选取的两类问题:1)选取k个领导者使系统误差达到最小;2)系统误差在一定范围内,选取最小数量的领导者;提出了一类超模博弈建模方法.在建模过程中设计了各个智能体的效用函数与系统整体的目标函数,使各个智能体在寻求各自效用函数最大化的过程中实现整体目标.而后,运用贪婪算法优化了智能体决策过程.本文分析了所建模型的平衡点存在性和系统稳定性.最后,利用仿真实例对比说明了本文提出的基于超模博弈的二阶多智能体系统领导者选择算法的有效性.     

Customer satisfaction‐aware scheduling for utility maximization on geo‐distributed data centers

With the increasingly growing amount of service requests from the world‐wide customers, the cloud systems are capable of providing services while meeting the customers' satisfaction. Recently, to achieve the better reliability and performance, the cloud systems have been largely depending on the geographically distributed data centers. Nevertheless, the dollar cost of service placement by service providers (SP) differ from the multiple regions. Accordingly, it is crucial to design a request dispatching and resource allocation algorithm to maximize net profit. The existing algorithms are either built upon energy‐efficient schemes alone, or multi‐type requests and customer satisfaction oblivious. They cannot be applied to multi‐type requests and customer satisfaction‐aware algorithm design with the objective of maximizing net profit. This paper proposes an ant‐colony optimization‐based algorithm for maximizing SP's net profit (AMP) on geographically distributed data centers with the consideration of customer satisfaction. First, using model of customer satisfaction, we formulate the utility (or net profit) maximization issue as an optimization problem under the constraints of customer satisfaction and data centers. Second, we analyze the complexity of the optimal requests dispatchment problem and rigidly prove that it is an NP‐complete problem. Third, to evaluate the proposed algorithm, we have conducted the comprehensive simulation and compared with the other state‐of‐the‐art algorithms. Also, we extend our work to consider the data center's power usage effectiveness. It has been shown that AMP maximizes SP net profit by dispatching service requests to the proper data centers and generating the appropriate amount of virtual machines to meet customer satisfaction. Moreover, we also demonstrate the effectiveness of our approach when it accommodates the impacts of dynamically arrived heavy workload, various evaporation rate and consideration of power usage effectiveness. Copyright © 2014 John Wiley & Sons, Ltd.     

New directions in distributed Nash equilibrium seeking based on monotone operator theory

There has been an increasing research interest in modeling, optimization and control of various multi-agent networks that have wide applications in industry, defense, security, and social areas, such as computing clusters, interconnected micro-grid systems, unmanned vessel swarms \cite{ChenJie}, power systems\cite{MeiS}, multiple UAV systems\cite{KolaricP} and sensor networks\cite{LiuR}. For non-cooperative agents that only concern selfish profit-maximizing, the decision making problem can be modelled and solved with the help of game theory, while Nash equilibrium (NE) seeking is at the core to solve the non-cooperative multi-agent games \cite{HenrikSandberg, IsraelAlvarez, Jong-ShiPang}. Distributed NE seeking methods are appealing compared with the center-based methods in large-scale networks due to its scalability, privacy protection, and adaptability. Recently, monotone operator theory is explored for distributed NE seeking, which is shown to provide an uniform framework for various algorithms in different scenarios. It has been gradually developing into a cutting-edge research field, with the prospect and necessity of future in-depth research. In non-cooperative multi-agent games, each agent has different characteristics and pursues maximizing its own benefit. Hence, there is no centralized manager that can force all agents to adopt specified strategies to optimize the overall benefits. Under the NE, no player can decrease its cost by unilaterally changing its local decision to another feasible one. To seek an NE, the agent is required to optimize its own objective function given the opponent''s countermeasures. Therefore, various optimization-based methods have been investigated for distributed NE seeking, such as the gradient flow method and the best response method....     

A new marketing strategy map for direct marketing

Direct marketing is one of the most effective marketing methods with an aim to maximize the customer’s lifetime value. Many cost-sensitive learning methods which identify valuable customers to maximize expected profit have been proposed. However, current cost-sensitive methods for profit maximization do not identify how to control the defection probability while maximizing total profits over the customer’s lifetime. Unfortunately, optimal marketing actions to maximize profits often perform poorly in minimizing the defection probability due to a conflict between these two objectives. In this paper, we propose the sequential decision making method for profit maximization under the given defection probability in direct marketing. We adopt a Reinforcement Learning algorithm to determine the sequential optimal marketing actions. With this finding, we design a marketing strategy map which helps a marketing manager identify sequential optimal campaigns and the shortest paths toward desirable states. Ultimately, this strategy leads to the ideal design for more effective campaigns.     

Understanding children's collaborative interactions in shared environments

  Traditional computer technology offers limited support for face-to-face, synchronous collaboration. Consequently, children who wish to collaborate while using computers must adapt their interactions to the single-user paradigm of most personal computers. Recent technological advances have enabled the development of co-located groupware systems offering support for concurrent, multi–user interactions around a shared display. These systems provide a unique collaboration environment in which users share both the physical and the virtual workspace. This paper examines how such technology impacts children's collaboration. Findings from this research show that when concurrent, multi–user interaction is supported on a shared display, children exhibit collaborative behaviour similar to their interactions during paper-based activities. The findings also suggest strengths and weaknesses of various mechanisms for supporting synchronous interactions that have implications for the design of computer systems to support children's face-to-face collaboration.     

Comparison procedure of predicting the time to default in behavioural scoring

The paper deals with the problem of predicting the time to default in credit behavioural scoring. This area opens a possibility of including a dynamic component in behavioural scoring modelling which enables making decisions related to limit, collection and recovery strategies, retention and attrition, as well as providing an insight into the profitability, pricing or term structure of the loan. In this paper, we compare survival analysis and neural networks in terms of modelling and results. The neural network architecture is designed such that its output is comparable to the survival analysis output. Six neural network models were created, one for each period of default. A radial basis neural network algorithm was used to test all six models. The survival model used a Cox modelling procedure. Further, different performance measures of all models were discussed since even in highly accurate scoring models, misclassification patterns appear. A systematic comparison ‘3 + 2 + 2’ procedure is suggested to find the most effective model for a bank. Additionally, the survival analysis model is compared to neural network models according to the relative importance of different variables in predicting the time to default. Although different models can have very similar performance measures they may consist of different variables. The dataset used for the research was collected from a Croatian bank and credit customers were observed during a 12-month period. The paper emphasizes the importance of conducting a detailed comparison procedure while selecting the best model that satisfies the users’ interest.     

Evolutionary algorithms for the multi‐objective test data generation problem

Automatic test data generation is a very popular domain in the field of search‐based software engineering. Traditionally, the main goal has been to maximize coverage. However, other objectives can be defined, such as the oracle cost, which is the cost of executing the entire test suite and the cost of checking the system behavior. Indeed, in very large software systems, the cost spent to test the system can be an issue, and then it makes sense by considering two conflicting objectives: maximizing the coverage and minimizing the oracle cost. This is what we did in this paper. We mainly compared two approaches to deal with the multi‐objective test data generation problem: a direct multi‐objective approach and a combination of a mono‐objective algorithm together with multi‐objective test case selection optimization. Concretely, in this work, we used four state‐of‐the‐art multi‐objective algorithms and two mono‐objective evolutionary algorithms followed by a multi‐objective test case selection based on Pareto efficiency. The experimental analysis compares these techniques on two different benchmarks. The first one is composed of 800 Java programs created through a program generator. The second benchmark is composed of 13 real programs extracted from the literature. In the direct multi‐objective approach, the results indicate that the oracle cost can be properly optimized; however, the full branch coverage of the system poses a great challenge. Regarding the mono‐objective algorithms, although they need a second phase of test case selection for reducing the oracle cost, they are very effective in maximizing the branch coverage. Copyright © 2011 John Wiley & Sons, Ltd.     

Extracting Actionable Knowledge from Decision Trees

Most data mining algorithms and tools stop at discovered customer models, producing distribution information on customer profiles. Such techniques, when applied to industrial problems such as customer relationship management (CRM), are useful in pointing out customers who are likely attritors and customers who are loyal, but they require human experts to postprocess the discovered knowledge manually. Most of the postprocessing techniques have been limited to producing visualization results and interestingness ranking, but they do not directly suggest actions that would lead to an increase in the objective function such as profit. In this paper, we present novel algorithms that suggest actions to change customers from an undesired status (such as attritors) to a desired one (such as loyal) while maximizing an objective function: the expected net profit. These algorithms can discover cost-effective actions to transform customers from undesirable classes to desirable ones. The approach we take integrates data mining and decision making tightly by formulating the decision making problems directly on top of the data mining results in a postprocessing step. To improve the effectiveness of the approach, we also present an ensemble of decision trees which is shown to be more robust when the training data changes. Empirical tests are conducted on both a realistic insurance application domain and UCI benchmark data     

A genetic algorithm approach for solving the daily photograph selection problem of the SPOT5 satellite

This article addresses the combinatorial optimization problem of managing earth observation satellites (EOSs) such as the French SPOT5, which is concerned with selecting on each day a subset of a set of candidate photographs. The problem has a significant economic importance due to its high initial investment cost that exists in these instruments and its solution difficulty resulting from the large solution space, making it an attractive research area. This article proposes a genetic algorithm (GA) for solving the SPOT5 selection problem using a new genome representation for maximizing not only a single objective as profit but a multi-criteria objective that includes the number of acquired photographs. Test results of our proposed GA show that it finds optimal solutions effectively for moderate size problems and obtains better results for two large benchmark instances coded 1403 and 1504 in the literature. Also, we verify the result that the best known value in the literature for problem coded 1401 is an optimal value.     

Monocular passive ranging of maneuvering intruder for UAS detect‐and‐avoid

We investigate a planar orbiting intruder passive ranging problem, where an ownship aircraft is moving with a constant velocity and a co‐planar intruder aircraft is conducting an orbiting maneuver. The objective is to estimate the state information of the intruder, such as range, speed and heading, using only bearing measurements of the intruder from the ownship. This is relevant to detect‐and‐avoid applications where an unmanned aerial system (UAS) is equipped with only a monocular camera sensor and needs to determine the range to the intruder. We utilize a filter bank algorithm parameterized with respect to the range, heading, and angular velocity of the intruder's initial states. In previous work, we performed investigations on the effects of parameterizing the filter bank algorithm and the system model using Monte Carlo Simulations. We extend our previous study to include a study on the effects of the ownship speed on the filter performance as well as a Cramer‐Rao lower bound comparison. Since the intruder's motion may be composed of straight legs and orbits, we integrate our filter bank algorithm into an Interacting Multiple Model (IMM) framework for two maneuver models: constant velocity and coordinated turn. We conduct Monte Carlo simulations of the IMM algorithm to study the effects of varying the IMM mixing stage frequency and the off diagonal values for the transition probability matrix. Our simulation results demonstrate the effectiveness of the proposed algorithms.     

NEGO — Group decision support system

There are two major frameworks for decision making: maximizing and satisficing. A combination of both may be used to describe group decision making (GDM). In the satisficing approach, decision makers (DMs) formulate aspiriation levels or demands which take the form of constraints. Choosing from among different decisions, DMs take into account their preferences or wants, which take the form of objective functions.GDM is divided into two stages: first, each DM makes a decision, and second, DMs negotiate so as to achieve a compromise decision. Negotiating is an iterative process. Negotiations are completed when all demands have been met.The group decision support system “NEGO” assists DMs in finding a compromise. It has been used for solving a GDM problem at the corporate level and is currently utilized in management courses.     

A hybrid memetic algorithm for maximizing the weighted number of just-in-time jobs on unrelated parallel machines

This paper presents a hybrid memetic algorithm for the problem of scheduling n jobs on m unrelated parallel machines with the objective of maximizing the weighted number of jobs that are completed exactly at their due dates. For each job, due date, weight, and the processing times on different machines are given. It has been shown that when the numbers of machines are a part of input, this problem is NP-hard in the strong sense. At first, the problem is formulated as an integer linear programming model. This model is practical to solve small-size problems. Afterward, a hybrid memetic algorithm is implemented which uses two heuristic algorithms as constructive algorithms, making initial population set. A data oriented mutation operator is implemented so as to facilitate memetic algorithm search process. Performance of all algorithms including heuristics (H1 and H2), hybrid genetic algorithm and hybrid memetic algorithm are evaluated through computational experiments which showed the capabilities of the proposed hybrid algorithm.     

Heuristics for a multiperiod inventory routing problem with production decisions

Manufacturers who resupply a large number of retailers on a periodic basis continually struggle with the question of how to formulate a replenishment strategy. This paper presents a comparative analysis of a series of heuristics for an inventory routing problem (IRP) that arises in a manufacturing supply chain. The IRP is formulated as a mixed integer program with the objective of maximizing the net benefits associated with making deliveries in a specific time period to a widely dispersed set of customers. It is assumed that inventory can accumulate at the customer sites, but that all demand must be met without backlogging. Because optimal solutions were not within reach of exact methods, a two-step procedure was developed that first estimates daily delivery quantities and then solves a vehicle routing problem for each day of the planning horizon. As part of the methodology, a linear program is used to determine which days it is necessary to make at least some deliveries to avoid stockouts.The IRP is investigated in the context of an integrated production–inventory–distribution–routing problem (PIDRP). The full model takes into account production decisions and inventory flow balance in each period. For the computations, a previously developed branch-and-price algorithm is used that requires the solution of multiple IRPs (one in each period) to generate columns for the master problem. Testing showed that PIDRP instances with up to eight time periods and 50 customers can be solved within 1 h. This level of performance could not be matched by either CPLEX or an exact version of the branch-and-price algorithm.     

基于分层结构的网络入侵检测模型的研究与实现

文章介绍了一种基于分层结构的网络入侵检测模型,它包括以下三部分:一个中心控制级检测模块,多个主机级检测模块和网络代理级检测模块。它们之间通过代理协同检测网络行为,实现实时入侵检测。另外,还介绍了各部分之间的通信机制。     

An integrated hierarchical programming approach for industrial planning

This paper describes of prototype of a microcomputer implementation of an integrated multicriteria expert support system (MCESS). The system is an interactive, comprehensive, and easy to use tool to support the manager in project selection and resource allocation. The MCESS combines the capabilities of goal programming, the analytic hierarchy process, net present value analysis, and a spreadsheet.

The literature on modeling with spreadsheets and on software integration is reviewed. Goal programming, a multicriteria decision making technique is described and the analytic hierarchy process is shown to be able to overcome some of its limitations. The structure of the MCESS is described. An illustration of its use in industrial planning is presented.     

Modeling machine loading problem of FMSs and its solution methodology using a hybrid tabu search and simulated annealing-based heuristic approach

In this article, a machine loading problem of a flexible manufacturing system (FMS) is discussed having the bicriterion objectives of minimizing system unbalance and maximizing throughput in the presence of technological constraints such as available machining time and tool slots. A generic 0–1 integer programming formulation with the objective functions and constraints described above has been proposed. A hybrid algorithm based on tabu search and simulated annealing (SA) is employed to solve the problem. The main advantage of this approach is that a short-term memory provided by the tabu list can be used to avoid revisiting the solution while preserving the stochastic nature of the SA method. The proposed methodology has been tested on ten standard problems and the results obtained are compared with those from some of the existing heuristics.     

An ant algorithm for optimization of hole-making operations

This paper deals with the optimization of hole-making operations in conditions where a hole may need several tools to get completed. The objective of interest in the considered problem is to minimize the summation of tool airtime and tool switch time. This objective is affected by the sequence through which each operation of each hole is done. The problem is formulated as a 0–1 non-linear mathematical model. An ant algorithm is developed to solve the proposed mathematical model. The paper includes an illustrative example which shows the application of the proposed algorithm to optimizing the sequence of hole-making operations in a typical industrial part. The performance of the proposed algorithm is tested through solving six benchmark problems. The computational experience conducted in this research indicates that the proposed method is both effective and efficient.     

PTRA:一个面向空载出租车的路线推荐算法

目前针对出租车的推荐系统主要从降低空载率和减少寻客里程两个方面为司机推荐潜在载客点或最优行驶路线,然而从司机收益最大化的角度而言,多数研究没有考虑实时路况的变化对推荐效果的影响.因此,以收益最大化为目标进行了研究,提出了一种空载出租车推荐算法PTRA(profit-based taxi recommendation a...     

Approximate algorithms for the container loading problem

In this paper we develop several algorithms for solving three–dimensional cutting/packing problems. We begin by proposing an adaptation of the approach proposed in Hifi and Ouafi (1997) for solving two–staged unconstrained two–dimensional cutting problems. We show how the algorithm can be polynomially solved for producing a constant approximation ratio. We then extend this algorithm for developing better approximate algorithms. By using hill–climbing strategies, we construct some heuristics which produce a good trade–off between the computational time and the solution quality. The performance of the proposed algorithms is evaluated on different problem instances of the literature, with different sizes and densities (a total of 144 problem instances).