Sequential Search and Selection Problem Under Uncertainty

This paper formulates and discusses a series of sequential decision problems of the following common structure: A decision alternative of multiple attributes‐that is, a job, an employee, or an investment alternative‐is to be selected within a certain fixed length of time. An unknown number of alternatives are presented sequentially, either deterministically or in a random manner. The decision maker can rank all the alternatives from best to worst without ties, and the decision to accept or reject an alternative is based solely on the relative ranks of those alternatives evaluated so far. The nonparametric sequential decision problem is first studied for a model involving a discrete time period and then generalized in terms of continuous time. Also considered is a variant of this problem involving a Bayesian estimation of (1) the uncertain probability of having an alternative at a given stage in the discrete‐time model and (2) the arrival rate of alternatives in the continuous‐time model. The optimal selection strategy that maximizes the probability of selecting the absolute best alternative is illustrated with the job search problem and the single‐machine job assignment problem.     

The sensitivity of job mix and load capacity bottlenecks on inventory and due date performance in a manufacturing system

Based on a benchmark job-lot manufacturing system a simulation study was carried out to compare the performance of just-in-time (JIT) shop control system Kanban with conventional job-shop control procedures. The shop control policies were tested under a good manufacturing environment and the effects of job mix and load capacity bottlenecks on various shop control policies were tested. From the simulation results, it is inferred that there are shop control procedures that perform better than the Kanban in a job shop. It has been observed that even with adequate capacity, bottleneck areas surface due to fluctuations in the shop load. Kanban is not appropriate in such a situation because capacity bottlenecks can significantly reduce the ell'ectiveness of a pull system. The disparateness in the processing requirements for jobs can seriously undermine the performance of the shop. This is the type of shop environment where the shop control procedures will be most effective. Although Kanban came out best when the load capacity bottlenecks and the disparateness of the job mix were removed, the selected shop control variable combinations closely approximated the Kanban result. Although many features of JIT can be implemented in any system, companies trying to adopt JIT should remember that Kanban requires a rigid system intolerant of any deviation.     

AHP Is Intended for More Than Expected Value Calculations

In a previous paper, Schoner and Wedley [9] presented a way to automatically compute the importance of criteria from the values of alternatives. The following discussion illustrates that the approach of [9] can only be applied if the alternatives under the criteria are measured with a single absolute scale. However, when this does occur, in part or in all of a decision problem, the approach is extraneous. The measurements can (and should) first be combined with the appropriate arithmetic of that scale which reduces that part of the problem to a single dimension and the result can be taken up with the remaining dimensions for consideration with AHP.     

Methods for comparing policies using multiple criteria: an urban example

This paper looks at a variety of methods that can be used in evaluating a set of alternate policies using multiple criteria. The methods examined are the structural mapping of indifferences, utility values, lexicographic ordering, factor analysis, concordance analysis and multidimension scaling. Each method is tested using hypothetical data for a problem in which alternative policies are proposed for allocating monies to housing and health projects in a town. The aim is to try to reveal as objectively as possible, a set of preferred alternatives from which one can be chosen in the political decision-making process. After describing and testing the methods individually, they are compared both on the basis of their results and on the principles involved in their approach. Conclusions about the validity of each method are given, and it is emphasized that all methods should only be used as aids in the choice of an optimal policy.     

The analytic hierarchy process with interval preference statements

In the analytic hierarchy process (AHP), interval judgments instead of precise ratios are widely accepted and can be practically used to solve decision-making problems when uncertainty exists because of scant information available or insufficient understanding of the problem. This paper presents a simple and effective method for finding the extreme points in a range of interval ratios (such as loose articulation, minimum number of interval ratios, and general interval ratios) and ultimately for establishing the dominance relations among alternatives using the identified extreme points. This is followed by an enumeration or simulation approach to manage situations in which the best or a full ranking of alternatives remains unidentified.     

OPTIMAL DESIGN OF MULTI-PRODUCT MULTI-ECHELON INVENTORY SYSTEMS*

Large scale inventory distribution systems typically comprise a hierarchy of retail stores and warehouses. This paper presents a model for finding the optimal design of such systems. Given the maximum number of facilities under consideration and their locations, the problem is to determine which facilities to include in the system and which products to stock at each in order to minimize the cost of the system. Demand for the products may be deterministic or stochastic. To use the model it is necessary to know the optimal inventory policies for the multi-echelon systems under consideration; however, an important feature of this work is that any multi-echelon model may be used in tandem with this design model. Included is an example to illustrate the model and the two points which are the basis for its formulation: first, there is generally no single design which is best for all products; second, the design which is optimal for a given product is not necessarily the best design to use for that product when trying to minimize the cost of the entire system.     

考虑产品变质的VMI混合补货发货策略及优化仿真

本文研究了“产品可变质”情况下的VMI库存补货与装运调度问题,并建立了Poisson需求过程下的VMI混合补货发货模型,根据此模型通过简单的规划求解即可得到使长期平均成本最小的最佳混合策略组合。由于模型推演过程中涉及到对补货周期内期望发货次数的近似估计,因而模型解是拟最优的。算例和模型仿真显示,模型结果与仿真结果十分相近,从而模型有效性得以确认。     

Social Security Wealth and Retirement Decisions in Italy

Abstract. This paper uses administrative data to study the retirement decisions of Italian private‐sector non‐agricultural employees during the period 1977–97. Our analysis tries to assess the importance of the financial incentives built into the social security system. The basic idea is very simple: at any given age, and based on the available information, workers compare the expected present value of two alternatives: retiring today or working one more year, and then choose the best one. A key role in this kind of comparisons is played by social security wealth, whose level and changes reflect the expectations about the profile of future earnings and the institutional features of the social security system. The various incentive measures that we consider differ in the precise weight given to the social security wealth that workers accrue as they continue to work. Our model does not provide a structural representation of the retirement process. A worker's decision is modeled here following a ‘quasi reduced‐form’ approach, with the incentive measures entering as predictors of the worker's choice in addition to standard variables. The estimated models are then used to predict retirement probabilities under alternative policies that change social security wealth and derived incentive measures.     

Dynamic Group Scheduling Heuristics in a Flow-through Cell Environment*

The increased use of cellular manufacturing configurations designed to grapple with increasing competitive pressures is providing manufacturing managers and engineers with a broad variety of operational challenges. Many questions concerning the best procedures and policies for the day-to-day operation of manufacturing cells are still unanswered. The primary objective of this study is to compare the performance of traditional single-stage heuristics and the two-stage group scheduling heuristics that have exhibited superior performance in previous studies in a flow-through cell environment under a rigorous set of experimental conditions. Such a comparison is of great interest since each previous study has focused on proposing new heuristics and testing them against some particular baseline heuristic, often without comprehensive comparisons to the broad variety of previously proposed heuristics. Two single-stage heuristics and four two-stage heuristics are examined under sixteen experimental conditions (four experimental factors at two levels each). The experimental factors examined are shop load, due date tightness, setup to run-time ratio, and interarrival time distribution. Results vary by experimental condition and performance criteria, but in general, two-stage heuristics outperformed single-stage heuristics under all experimental conditions, as well as being relatively insensitive to changing experimental conditions. In addition, two of the two-stage heuristics displayed superior performance on all performance measures under most experimental conditions. Finally, the results indicated that interarrival time distribution does have a major impact on the performance of scheduling heuristics.     

Optimal Selection Strategy for the Group Interview Problem

We propose a more generalized version of the secretary problem, called the group interview problem, in which each group contains several alternatives and each group of alternatives is presented and evaluated sequentially over time. Using the assumptions corresponding to the classical secretary problem, we derive an optimal selection strategy which maximizes the probability of winning or selecting the single best choice in a given sequence of groups. We further address the problem of choosing at the beginning of the evaluation process a sequence of groups to maximize the winning probability. Because of formidable computational requirements to obtain an optimal solution to this sequencing problem, we then develop a heuristic algorithm based on several properties inherent in an optimal selection strategy. The heuristic procedure is evaluated experimentally using Monte Carlo simulation and is shown to be effective in obtaining near-optimal (within 5 percent) solutions.     

RUTA: A framework for assessing and selecting additive value functions on the basis of rank related requirements

In this paper, we present a new preference disaggregation method, called RUTA, which infers a set of additive value functions from the preference information referring to the desired ranks of some reference alternatives. Real-life experience indicates that people willingly refer to the range of allowed ranks that a particular alternative should attain, or to constraints on the final scores of the alternatives. We develop a mathematical model for incorporating such preference information via mixed-integer linear programming (MILP). Then, we discuss how decision making could be supported with the use of the already proposed extreme ranking analysis (ERA), which indicates the best and worst ranks gained by each alternative over the set of compatible preference model instances. We also introduce a new interactive UTA-like technique, which aims at selecting a single value function representing the outcomes of ERA. In the interactive process, the decision maker (DM) is assigning priorities to different pre-defined targets, which are built on results of ERA, and refer to the comparison of the best and/or worst ranks for pairs of alternatives. In particular, the DM may choose to emphasize or neglect the advantage of some alternatives over the others, in terms of results of ERA. In this way, one obtains a synthetic representation of extreme ranking analysis at a higher level of abstraction.     

Supply Performance in Multi‐Echelon Inventory Systems with Intermediate Product Demand: A Perspective on Allocation

In this article, we study the performance of multi‐echelon inventory systems with intermediate, external product demand in one or more upper echelons. This type of problem is of general interest in inventory theory and of particular importance in supply chain systems with both end‐product demand and spare parts (subassemblies) demand. The multi‐echelon inventory system considered here is a combination of assembly and serial stages with direct demand from more than one node. The aspect of multiple sources of demands leads to interesting inventory allocation problems. The demand and capacity at each node are considered stochastic in nature. A fixed supply and manufacturing lead time is used between the stages. We develop mathematical models for these multi‐echelon systems, which describe the inventory dynamics and allow simulation of the system. A simulation‐based inventory optimization approach is developed to search for the best base‐stock levels for these systems. The gradient estimation technique of perturbation analysis is used to derive sample‐path estimators. We consider four allocation schemes: lexicographic with priority to intermediate demand, lexiographic with priority to downstream demand, predetermined proportional allocation, and proportional allocation. Based on the numerical results we find that no single allocation policy is appropriate under all conditions. Depending on the combinations of variability and utilization we identify conditions under which use of certain allocation polices across the supply chain result in lower costs. Further, we determine how selection of an inappropriate allocation policy in the presence of scarce on‐hand inventory could result in downstream nodes facing acute shortages. Consequently we provide insight on why good allocation policies work well under differing sets of operating conditions.     

Approximate weighting method for multiattribute decision problems with imprecise parameters

A multiattribute decision problem with imprecise parameters refers to one in which at least one of the parameters such as attribute weights and value scores is not represented by precise numerical values. Some well-known types of incomplete attribute weights are chosen and analyzed to find their extreme points. In doing so, we show that their coefficients matrix, by itself or by the change of variables, belongs to a class of M-matrix which enables us to find its extreme points readily due to the inverse-positive property.The knowledge of extreme points not only helps us to prioritize alternatives but also supports iterative exploration of decision-maker’s preference by investigating modified extreme points caused by additional preference information. A wide range of eligible attribute weights, however, often fail to result in the best alternative or a complete ranking of alternatives. To address this situation, we consider an approximate weighting method, so called the minimizing squared deviations from extreme points (MSD) which locates the attribute weights at the barycenter of a weight set. Accordingly, the MSD approach extends the rank order centroid (ROC) weighting method which is known to outperform other approximate weighting methods in case of ranked attribute weights. The evidence of the MSD’s superiority over a linear program-based weighting method is verified via simulation analysis under different forms of incomplete attribute weights.     

RANK GENERATION,PRESERVATION, AND REVERSAL IN THE ANALYTIC HIERARCHY DECISION PROCESS

Decision making has the objective of finding the best alternative or set of alternatives by considering a number of goals, objectives, criteria, competitors, and other important factors. The analytic hierarchy process is a decision aid used to assist a decision maker in sorting out the complexity of a decision problem and making use of his or her judgments. A decision maker must be assured that the arithmetic operations of any such decision process are the right ones—that they surface the correct ranking and values of the alternatives and preserve or alter ranks appropriately when new alternatives are added or deleted. In this paper it will be shown that with absolute measurement, rank always is preserved, with relative measurement, rank changes with nspect to scveral criteria only because of the structural dependence (involving both numbers and measurements) of criteria on alternatives. A discussion of the effect on rank of replicas and near replicas of the alternatives also is given.     

The coming alliance revolution in health care

Like it or not, the health care profession is being "shifted" into a revolutionary new world. The question is not will it change but rather how will it change? Who will determine its fate? What form will these changes take? What are the best alternatives for physicians, institutions, health care workers, insurers, employers, and, most importantly, patients? Some of the changes will come from government mandate, others from market forces. To understand what the future might bring, we should look at both the driving forces behind the changes and how other industries have responded to similar forces. An important consideration for health care professionals will be how, if at all, the concepts of collaboration and cooperation that are inherent in networking and alliances will guide their planning.     

Tools for Comparative Analysis of Alternatives: Competing or Complementary Perspectives?

A third generation of environmental policy making and risk management will increasingly impose environmental measures, which may give rise to analyzing countervailing risks. Therefore, a comprehensive analysis of all risks associated with the decision alternatives will aid decision-makers in prioritizing alternatives that effectively reduce both target and countervailing risks. Starting with the metaphor of the ripples caused by a stone that is thrown into a pond, we identify 10 types of ripples that symbolize, in our case, risks that deserve closer examination: direct, upstream, downstream, accidental risks, occupational risks, risks due to offsetting behavior, change in disposable income, macro-economic changes, depletion of natural resources, and risks to the manmade environment. Tools to analyze these risks were developed independently and recently have been applied to overlapping fields of application. This suggests that either the tools should be linked in a unified framework for comparative analysis or that the appropriate field of application for single tools should be better understood. The goals of this article are to create a better foundation for the understanding of the nature and coverage of available tools and to identify the remaining gaps. None of the tools is designed to deal with all 10 types of risk. Provided data suggest that, of the 10 types of identified risks, those associated with changes in disposable income may be particularly significant when decision alternatives differ with respect to their effects on disposable income. Finally, the present analysis was limited to analytical questions and did not capture the important role of the decision-making process itself.     

Rationality in policy making: Two concepts not one

The concept of rationality, central to any analysis of policy making, is shown to be dual. The qualitative concept of rationality, broadly equivalent to ‘appropriate’ directs the policy analyst to an examination of information flow (since it may be defective), of goals (since these may conflict), and of the momentum of decisions (since it can produce irrationality). The quantitative concept of rationality, broadly equivalent to ‘efficient’ is itself divisible into two models (‘ends–means’ and ‘alternative behaviour’). Either, however, directs the policy analyst to an exploration of alternatives (since ideally all should be known) and of evaluative techniques (since ideally these should indicate the best alternative). Quantitative, but not qualitative, rationality is frontally challenged by some theorists (notably Lindblom) and claimed to be insufficient by others (notably Dror).     

Development and analysis of mathematical and simulation models of decision-making tools for remanufacturing

This article presents decision-making tools for remanufacturing. The first decision-making tool was used to address inventory lot-sizing problems in a hybrid remanufacturing–manufacturing system with varying remanufacturing fraction. In this article, the new inventory lot-sizing model with variable remanufacturing lot sizes has been shown to exhibit better performance than the benchmark model with fixed remanufacturing lot sizes. The new inventory lot-sizing model is anticipated to become a valuable decision-making tool in companies that are planning to adopt remanufacturing. The second decision-making tool was applied to address a production and inventory planning problem in a remanufacturing system considering different remanufacturing policies for a given remanufacturing strategy. For a remanufacture-to-stock system with two quality remanufacturables groups four alternative policies were examined, a policy which specifies simultaneous processing utilising dedicated resources was shown to be the best policy to achieve a shorter remanufacturing cycle time. For a remanufacture-to-order system with two quality remanufacturables groups, the three relevant policies of the four alternative policies were examined, a policy which specifies sequential processing and switching between various quality remanufacturables groups was shown to be the best policy to achieve a shorter remanufacturing cycle time. The production and inventory planning simulation models in a remanufacturing system are expected to become significant decision-making tools in remanufacturing operations.     

A Probabilistic Analysis of the Passive–Restraint Question

The benefits and costs of automobile safety policies are compared using a methodology which explicitly quantifies the uncertainties. The policies addressed include both voluntary and compulsory manual belt usage, nondetachable passive (automatic) seat belts, and air bags. Estimates of the effectiveness and usage rates of these alternatives were obtained in the form of subjective probability distributions from eight experts. Their opinions were combined using equal weighting. The direct economic costs of the technologies were also estimated probabilistically. The number of lives saved and the net benefits of the policies were calculated probabilistically for a range of values of lifesaving. Probabilistic computations and sensitivity analysis were performed by the Demos modelling system using Monte Carlo simulation. The results are highly uncertain and quite sensitive to the value of lifesaving. Nevertheless, they imply that repeal of the passive-restraint standard is defensible according to the net-benefit criterion only if a relatively low value is assigned to lifesaving. The degree of uncertainty emphasizes the potential value of demonstration programs to obtain better information.