Data Driven Modeling: An Application in Scheduling

This paper presents a data driven modeling (DDM) approach to certain types of optimization problems. DDM relinquishes control of the completed model to the user department rather than the operations research (OR) staff. The approach emphasizes development of models that are dependent on data maintained and understood by the users. The data base consists of coded user rules which describe when changes will occur in the problem structure and data which captures the generalization of the problem. Both the rules and data can be updated by user department personnel. These data drive a matrix generator controlled by the rules which uses the data base as input to generate the specific model formulation. This DDM system is designed by OR consultants or staff to allow independence of use along with low-cost and minimal-effort maintenance. The DDM approach is illustrated with an application to a real-world medical scheduling problem.     

A Multiperiod Audit Staff Planning Model Using Multiple Objectives: Development and Evaluation*

The audit staff planning problem, a specific type of manpower planning problem, has been modeled using goal programming and, more recently, multiple objective linear programming. Prior studies developed single-period models and did not go beyond the model building stage. This study develops a multiperiod audit staff planning model and evaluates the model using a test application involving actual decision makers (partners in public accounting firms). The multiperiod model includes seven objectives to be optimized: profit (to be maximized), late completion of work, work declined, staff augmentation, staff reduction, underutilization of the work force, and shortfall in meeting professional development targets (all to be minimized). Over a four-quarter planning horizon with one “busy season,” the model is subject to constraints with respect to the projected audit work load, ability to substitute personnel and to perform interim audit work, available staff hours (including overtime limitations), supervisory requirements, and professional development targets. Results of the test application showed that the model was capable of producing a range of values for each objective. The participants were exposed to much of that range when making their decisions. The results also showed that all objectives were important and that participants were consistent in choosing their preferred level of each objective over several runs of the model. These results and the reactions of the participants demonstrate that the model is usable by actual decision makers and has potential for a number of specific applications.     

Changing Patterns of Residential Care

The use of soft-tie restraint, a common means of controlling the assaultive behavior of mentally retarded residents in institutions, was examined with a retrospective design. Naturally occurring changes over a 39-month period were divided into (a) changes in management personnel that were associated with changes in management support of behavioral programming, (b) changes in staff training and feedback, and (c) a change in the physical location of the Program. Use of restraint decreased when the use of behavioral programming was supported by management and increased when it was not. When a four-person Training Team provided staff with training and formal and informal fecdback, restraint was used less than it was when the Training Team was outside the Program and training and feedback were not provided. Changing the physical location of the Program had little effect on the use of restraint. Relative to baseline, use of restraint decreased by 80% when management supported the use of behavioral programming and the Training Team was in the Program providing training and feedback. Reductions in the use of restraint were correlated with a corresponding reduction in Special Incident Reports which were filed when an event such as injury to a resident or staff member occurred.     

A Further Evaluation of Response Cards: Teaching Direct Care Staff Basic Behavioral Principles

Numerous studies have demonstrated the efficacy of response cards for teaching students in educational settings; however, the effectiveness of response cards has yet to be examined in organizational settings where such strategies may be very useful (i.e., in staff training). The purpose of this study was to evaluate the utility of response cards during in-service training to teach basic behavioral principles to direct care staff working in a human service setting. Results indicated that individuals who received the response card training participated more during training and were able to retain more information over time relative to those who received a standard in-service training.     

Visualizing association rules in hierarchical groups

Association rule mining is one of the most popular data mining methods. However, mining association rules often results in a very large number of found rules, leaving the analyst with the task to go through all the rules and discover interesting ones. Sifting manually through large sets of rules is time consuming and strenuous. Although visualization has a long history of making large amounts of data better accessible using techniques like selecting and zooming, most association rule visualization techniques are still falling short when it comes to large numbers of rules. In this paper we introduce a new interactive visualization method, the grouped matrix representation, which allows to intuitively explore and interpret highly complex scenarios. We demonstrate how the method can be used to analyze large sets of association rules using the R software for statistical computing, and provide examples from the implementation in the R-package arulesViz.     

AN OPTIMIZATION-BASED METHODOLOGY FOR RELEASE SCHEDULING

Operational-level scheduling decisions on the factory floor include the release policy and the dispatching rule. Recent work by researchers suggests that factory performance depends heavily on the release policy. Several heuristic factory release rules have been developed that generate release decisions based on the status of the bottleneck machine. The heuristic rules perform very well in reducing the work-in-process inventory. However, factory output schedules generated by these rules generally fail to match the demand pattern, creating excessive finished goods inventory and/or excessive backorders. To overcome these difficulties, a computationally efficient integer programming model is proposed for release scheduling.     

Annualized hours as a capacity planning tool in make-to-order or assemble-to-order environment: An agricultural implements company case

Capacity planning is one of the most important elements for an efficient production planning and control system. It is even more important when the main strategies to fulfil sales requirements are make-to-order (MTO) or assemble-to-order (ATO). Neither MTO or ATO companies keep finished products in stock which means that production level is driven by actual sales figures. Therefore, it is very important to have an available capacity as flexible as possible to meet sales requirements. Annualized hours (AH) is a work time control system that helps in increasing the flexibility of available capacity. The main rule of the AH system is to hire employees to work a certain number of hours on a yearly basis. Following some other pre-defined rules and limits, the AH system allows a company to adjust weekly available time in order to reduce idle capacity or to fulfil requirements that could mean overtime. This paper proposes a linear programming model to plan the operations using AH. This model has been implemented in a company that produces agricultural implements, showing that it is very simple to use. It is producing excellent practical results.     

银行个人贷款审批规则库的知识求精

鉴于个人贷款在银行业务中的比重不断上升,对个人贷款申请进行审批显得格外重要。由于初始规则库常常存在知识不完全或不一致等问题,所以需要规则库修正,即规则库知识求精。现以日本银行个人贷款审批规则库为例,采用神经网络技术,利用个人贷款审批的历史数据,对规则库进行知识求精。结果表明,求精后的规则数量减少了,而分类准确率却有很大的提高。     

飞行员模拟机复训计划优化模型研究

根据民航有关规定,飞行员每6个日历月需要参加一次模拟机复训.公司制定模拟机复训计划,要考虑模拟机资源、教员资源、生产任务允许模拟机复训的人数和飞行员训练适宜期,本文以机长和副驾驶配组数最大化和训练/检查月内参加复训人数最大化为目标,在此基础上建立了多目标规划问题,仅考虑配组目标时,给出了不存在隔月配组的条件.根据不同的目标,本文研究对象可分解为两个经典网络规划子问题,文中给出了求解的方法,并给出了一个实际案例.     

A simulation study of dispatch rules for reducing flow times in semiconductor wafer fabrication

Semiconductor wafer fabrication involves very complex process routing, and reducing flow times is very important. This study reports a search for better dispatch rules for achieving the goal of reducing flow times, while maintaining high machine utilization. We explored a new simulation-based dispatch rule and a queue prediction dispatch rule. Using simulation experiments and an industrial data set, we also compared several other dispatch rules commonly used in semiconductor manufacturing with our proposed dispatch rules. Among these rules, in addition to the simulation-based dispatching rule, the shortest-remaining-processing-time, earliest-due-date and leastslack rules also performed well in terms of reducing flow times. The reasons behind these good rules are discussed in this paper. Based on the previous works and this study, accurately predicting and effectively utilizing future flow times can improve the quality of production management decisions.     

Admission,Tuition, and Financial Aid Policies in the Market for Higher Education

We present an equilibrium model of the market for higher education. Our model simultaneously predicts student selection into institutions of higher education, financial aid, educational expenditures, and educational outcomes. We show that the model gives rise to a strict hierarchy of colleges that differ by the educational quality provided to the students. We also develop a new estimation procedure that exploits the observed variation in prices within colleges. Identification is based on variation in endowments and technology. It does not rely on observed variation in potentially endogenous characteristics of colleges such as peer quality measures and expenditures. We estimate the structural parameters using data collected by the National Center for Education Statistics and aggregate data from Peterson's and the National Science Foundation.     

基于云模型关联规则的企业转型战略风险预警

针对评价指标数据的特点,构造了一种基于云模型的数值型关联规则挖掘算法,并将其运用于企业转型战略风险预警。首先运用云模型约简评价指标;然后,采用属性空间软划分方法对定量型属性的定义域进行划分,使定量型关联规则挖掘转换为定性关联规则挖掘,此基础上提取规则模版;最后采用有规则约束的Apriori算法挖掘云关联规则,并对检验样本风险等级进行判别。实证分析结果表明,与标准BP神经网络模型相比,该模型是一种更为有效和实用的战略风险预警工具。     

Qualifikation von Ausbildungssupervisoren in der Sozialen Arbeit – Kontexte,Methoden und Erfahrungen

This article moots the issue of how to support students of social work and social pedagogy to develop the prerequisites for a professional exercise within their future vocational role by dint of supervision. Academical training has to offer spaces for reflecting the relationship between clients and their assistants and for the student’s self-development as well. Methodical settings are reflected for the vocational training of students. Of what kind are the relationships between knowledge transfer, practical guidance and supervision? This article pleads for supervision as an educational program and specifies the qualifications for supervisors.     

基于领域知识和聚类的关联规则深层知识发现研究

本文针对传统关联规则挖掘算法产生大量冗余规则,提出了对关联规则结果进行二次挖掘,并设计了算法对挖掘出的关联规则进行聚类,然后基于已有领域知识对聚类后的关联规则进行新颖度评价,对于新颖度较高价值较大的关联规则可以存储于领域知识库用于决策使用或再次挖掘过程。该算法有效的减少的规则的数量,提高了规则的新颖性和精确度,对商业应用具有很高的价值。文章最后使用UCI开源数据进行了实验分析,并验证了该算法的有效性。     

A comparison of dispatching rules for executing a resource-constrained project with estimated activity durations

This research examines the performance of 13 dispatching rules for executing a resource-constrained project whose estimated activity durations may differ from the actual activity durations. The dispatching rules are tested in environments characterized by three factors, namely, the order strength of the precedence relationship, the level of resource availability and the level of estimation errors in the activity durations. The results show that project environment affects only the performance differences but not the grouping of the better dispatching rules. The greatest number of successors, rank positional weight, greatest cumulative resource requirement and minimum activity slack dispatching rules consistently perform better than the other dispatching rules, unaffected by the accuracy of the estimated activity durations. This finding validates the results of many past studies in the deterministic project environment for choosing the right dispatching rule for both projects with and without well-estimated activity durations.     

基于意外度的关联规则深层知识发现及应用研究

为了弥补传统关联规则挖掘产生大量冗余规则、难以直接用于决策支持的不足,本文提出了一种基于用户已有知识的规则意外度评价方法,并在此基础上设计了基于意外度的深层关联规则挖掘算法。算法的优点在于能够将用户已知的规则作为领域知识加入到数据挖掘过程从而有效过滤和已知规则相近的冗余规则,并且可以将新得到的规则加入知识库中实现知识的积累和重用。最后本文采用一个商场数据验证了该算法的有效性,并且对具有回馈模式的关联规则在商品促销中的作用进行了分析。     

基于晶格场理论和动态规划的国债期货定价研究

通过用一个二维量子场取代传统金融上的布朗运动,构建了可有效纳入国债远期利率在到期时间和日历时间两个维度上的不完全相关性的量子场理论模型,并离散化二维量子场,得到国债远期利率的晶格场理论模型,同时结合动态规划方法,将国债期货的所有交易交割规则纳入一个模型进行建模,实现在统一的模型框架下对国债期货及其内嵌的择时期权和质量期权进行定价。研究结果亦表明,所构建的国债期货定价模型的定价效果均显著优于传统的主流两因子HJM模型,且与真实市场结算价的贴合性均很强,特别地,在临近交割月份,其定价误差均降至0.05%以内。而各国债期货合约的质量期权价值都在其对应的国债期货面值的2%至6%之间,其择时期权价值大部分时间都在0附近徘徊,但随交割月份临近,择时期权价值开始迅速上升,最大时约为期货合约面值的0.6%。     

Work force scheduling with the search decision rule

This paper presents the results of a simulation study of the application of the search decision rule to work force scheduling for a manufacturing firm. The problem is of a mixed structure which includes fixed charge components (hiring and training costs), integer variables (work force size in each period), and interacting variables (work force changes, hiring and training costs, and overtime costs). With these elements present, the search decision rule produced lower total costs than the actual company decisions had over the simulated time period. The technique was found to be (1) flexible, (2) easily understood and implemented, and (3) computationally efficient, requiring only a few seconds of computer time per run.     

Deliberate simple rule creation and use: Activities and challenges

Using ‘simple rules’ may enable managers to take organizational decisions more rapidly. While prior research presents advantages of simple rule use during strategy formation, we lack insights into how firms can deliberately create simple rules and mitigate the challenges during strategy implementation. This is particularly interesting for established firms struggling to leverage their wealth of experience. We explore how managers of a multinational corporation deliberately create and use simple rules to implement the firm's growth strategy. Drawing on interviews and secondary data, we reveal the activities through which managers ensure the relevance and legitimacy of codified simple rules, yet also establish causality between simple rules and outcomes. Simple rule creation is accomplished via bottom-up identification and lateral validation, its use via consistent top-down guiding and timely adaptation. Our findings contribute to the growing body of research on the evolution of simple rules and aspects of strategy implementation more generally.     

The Use of In Situ Behavioral Skills Training to Improve Staff Implementation of the PEAK Relational Training System

The study evaluated the effectiveness of an in situ behavioral skills training program for training 3 direct care staff to implement the Promoting the Emergence of Advanced Knowledge Relational Training System (PEAK) with 3 individuals with autism. PEAK is an evidence-based behavior analytic language development curriculum that uses discrete trials to promote the emergence of language. The data suggest that behavioral skills training improved staff implementation of PEAK and resulted in a corresponding improvement in selected language skills across 2 of 3 learners with autism.