1、浙江万里学院现代物流学 院 外文文献译文 1 毕业论文外文翻译 译文 标题：库存管理中知识发现系统的设计与开发 资料来源 : 新加坡南洋理工大学 作者： C. A. Mitrea, C. K. M. Lee 摘要：为了有效的管理库存，必须要有准确的预测。对与预测相关的知识的提取和部署吸引了学术界和实际工作者的关注。知识被认为是对企业有价值的资产，它可以被智能技术如人工神经网络（ ANN）所操纵。人工神经网络具有特殊的学习能力，能从观测中通过输入数据来获取一个知识领域。这项研究的重点在于探究人工神经网络如何学习，分析不同类型的人工神经网络和人工神经网络架构在需求预测中的应用。需求预测问题所提出的
2、方法的可行性是和数字数据一起展示的。这项研究的意义是采取人工神经网络作为知识发现系统从而提高库存管理。 关键字：预测，库存管理，知识发现过程，神经网络，规则提取，权重 1.引言 库存管理效率受预测准确性的影响很大，如果有关它的变量知识是可用的则效率可以加强。知识有一种容易被人们理解的形式，就是规则。系统是否有效率取决于系统规则的质量。然而，知识的获取是不容易的。在库存管理方面，有些知识不容易被知识工程师所描述，有些不容易被管理者理解还有一些不容易被智能系统 确定，这已经成为知识管理的瓶颈。人工神经网络被视为机器学习的工具，具有自我学习的能力，可以通过归纳机器学习的过程获得自己的规则。然而，通过
3、神经网络获得的规则是隐藏在网络的体系结构和权重中的，不容易被人们理解或用于改善库存管理。 神经网络学习的两个重要标准是对学习模型和预测准确性的理解。虽然神经网络能达到较高的预测准确性，但其预测行为的不可理解性阻止了它在解决库存管理问题中的应用。出于这个原因，这项研究是探索提取规则的技术和方法，嵌入一个神经网络的权重，这个神经网络被训练来预测需求和库存管理的交货时间。 2.文献回顾 本节提供了库存管理、知识发现过程和人工神经网络三个综合研究主题。库存管理需要详细规划，控制和监控商品与服务的流动和存储。为了实现令人满意的服务水平隐含的和隐形知识是必需的，以便加强库存管理。 2.1 库存管理是管理库
4、存的艺术，它被定义为是公司或企业持有的有意出售或将其转化为更有价值状态的物质资源。在过去的 15 年里，库存管理已经成为浙江万里学院现代物流学 院 外文文献译文 2 企业和组织竞争优势的一个重要焦点。 根据 Sprague 的研究 ,库存管理在企业中起着至关重要的作用，它性能的提高可以加强整体供应链活动。 Liberman 定义了五个决定库存管理 绩效的因素和企业面临的挑战。（ 1）技术因素，企业缺乏可用性，可靠性和高效率的知识技术；（ 2）组织因素，公司可能没有正确的技术投入和财政支持；（ 3）财政因素，公司可能没有明确的财务机制；（ 4）管理因素，公司没有足够的训练和适当的管理；（ 5）信
5、息因素，公司缺乏适当的信息和有关信息共享问题。另一个研究影响库存管理绩效因素的是 Vries，他把影响库存管理绩效的因素分组如下（ 1）经济，（ 2）行为和（ 3）组织类别。相同，在 Rajeev 的研究中，他确定影响因素如下（ 1）安全库存，（ 2）产能利用率水平，（ 3）采购成效，（ 4）需求预测的准确性，（ 5）库存管理的实践，（ 6）管理态度，（ 7）员工培训，（ 8）与供应商的互动，（ 9）与客户的互动和（ 10）供应商授权。 其他研究人员缩小了库存管理绩效技术方面的因素只研究预测的影响因素。安全库存是对库存的额外单位进行保护，防止可能的缺货。当使用订单报价和计算生产计划时，必须使用
6、安全库存，以防止不确定性的制造工艺，需求和交货时间。有些因素，如交货时间，客户数量，客户满意度，交货的可靠性，供应可靠性会影响安全库存。 Dongxu 确定的因素有使用频率，质量等级和实际安全库存，而 Jia 提议促进因素如 断货成本，销售形势。图 1 整合了决定安全库存的所有影响因素。 并非所有影响安全库存的因素都有相同的相关比重。如交货时间因素可能有一个更大的决定安全库存的幅度。要确定每个因素对神经网络预测模型的意义是这项研究的目标之一。 图 1 影响安全库存的因素 3.方法 本节提出了知识发现模型（图 4）和使用一组给定数据的计算例子。表 1是提取至 48 个总值中的前 9 个值。 客户
7、数量 实际安全库存 供应可靠性 客户满意度 影响安全库存的因素 交货可靠性 交货时间 产品质量和可靠性 存货周转率 浙江万里学院现代物流学 院 外文文献译文 3 表 1 太浩盐业数据 图 4 知识发现模型 知识发现模型的使用由九个步骤组成。本文只着重前七个步骤，规则提取的步骤将在今后的工作中进行深入分析。 分解方法基于一个提取算法，在算法中每个网络组件都被检查过。在这个层次中提取的知识随后合并。由于这种方法集中的各个组件，被认为是一个openbox 方法。 教育学（输入 /输出）方法集中在分析网络的输入 /输出行为。从本质上讲，该方法涉及到黑盒子。 这个实验的目的是证明在输入变量 和输出变量存
8、在关系的情况下，神经网络学习它们之间关系的事实。因此，在尝试了不同的神经网络体系结构后，时期（月） 1 2 3 4 5 6 7 8 9 季节因素 0.47 0.68 1.17 1.67 0.47 0.68 1.17 1.67 0.47 水平 18439 19015.4 19644.2 20325.4 21059 21845 22683.4 23574.2 24517.4 趋势 524 524 524 524 524 524 524 524 524 需求 8000 13251.16 24087.26 36306.13 10784.34 16458.04 29850.21 44881.92 132
9、96.39 随机需求 21234 34345 9643 44321 53423 12321 16897 33873 43212 （ 1）目标定义 （ 2）数据采集 （ 3）数据选择 （ 4）数据预处理 （ 5）数据转换 （ 6）神经网络体系结构的选择 NARX MFFN (7)性能分析 （ 8）规则提取 分解方法 教学方法 （ 9）使用知识 库存管理 浙江万里学院现代物流学 院 外文文献译文 4 NARX 网络被选择，它有 17 个在隐藏层的神经元和四个作为输入的变量：时期，季节因素，水平，趋势和需求作为输出（表 1）。假设一下，假设输入变量与输出安全库存（ S）之间存在关系，用方程描述， W
10、inter 方程： F t+l =(L t +lT t )S t+l， F t+n 是时期 l的预测需求， L t 是水平 , T t 是趋势， L t+1=(D t+1/S t+1)+(1-)(L t +T t) T t+1= (L t+1 -L t)+(1-)T t S t+p+1 =(D t+1 /L t+1)+(1-)S t+1 为了简化模型，我们假设 T 和 S 是不变的，需求的周期是 4。 例如，在第 2 阶段的预测需求的计算公式是： F2 =(L1 +2T 1)S 2=(18439+2*524)0.68=13251.16 因此，在测试网络时，输入层有 4 个变量， L1， T1，
11、 l 和 S2，同时输出是F2。 该网络使用默认的 Levenberg-Marquardt 算法来培训。应用程序把输入向量和目标向量分为三组： 60%被用来培训， 20%被用来 验证网络的一般化和在过度拟合之前停止培训，最后的 20%被用来作为网络泛化的完全独立测试。 6.结论 本文采用计算的例子表明，神经网络能够探测到隐藏的知识。除了检测嵌入在神经网络中的知识，神经网络可用于成功发现输入参数之间的关系。然而，知识的提取和用人们能够理解的方式表示还需要进一步调查。因此，进一步的研究是探讨如何表示库存管理的相关知识和调查在这项研究中是分解方法还是教学方法更合适。 神经网络模型在提高预测的准确性中
12、有了可喜的成果，无论是在时间序列还是因果模型上。提取嵌入在神经网络权重中的知识可以帮助了解相 关输入参数，这些因素确定了库存管理的效率。知识发现过程是从培训过的神经网络中雇佣和提取知识的过程。在本文中，知识发现过程的算法已经通过研究负责收集的步骤和知识处理与部署进行分析。 浙江万里学院现代物流学院 外文文献原文 2 外文文献原文 Title: Design and Development of a Knowledge Discovery System in Inventory Management Material Source: Nanyang Technological Universit
13、y, Singapore Author: C. A. Mitrea, C. K. M. Lee Abstract： To manage the inventory efficiently, it is necessary to have accurate forecasting. To extract and deploy the knowledge associated with forecasting attracts the attention of both academic and practitioners. Knowledge is regarded as a valuable
14、asset for enterprises and it can be manipulated through intelligence techniques like Artificial Neural Networks (ANN). ANN has the special ability to learn facts about one knowledge domain by inputting data obtained from observations. This study focuses on exploring how ANN learns and analyzes diffe
15、rent types of ANN and ANN architectures used in the demand forecasting. The feasibility of the proposed approach to the demand forecasting issue is demonstrated with numeric data. The significance of this study is to adopt ANN as a knowledge discovery system thereby enhancing the inventory managemen
16、t. Keywords： Forecasting, Inventory Management, Knowledge Discovery Process, Neural Networks, Rule Extraction, Weights 1. INTRODUCTION Inventory management efficiency is affected considerably by the forecasting accuracy which can be enhanced if knowledge about its variables is available. A form of k
17、nowledge which is easy to understand by humans is rules. Whether the system is efficient depends on the quality of rules in the system. However, knowledge acquisition is not easy. In the inventory management field, some knowledge cannot be described easily by knowledge engineers, some cannot be unde
18、rstandable easily by managers, and some cannot be identified easily by intelligent systems. This has been already the bottleneck of knowledge management. Considered as a tool of machine learning, an artificial neural network has the capability to self learning hence acquiring rules by itself through
19、 the process of inductive machine learning. However, rules obtained by neural networks are hidden in the architecture and weights of networks, and are not easy to understand by human being or to be applied to improve inventory management . Two crucial criteria for NN learning are based on comprehens
20、ibility of learned models and predictive accuracy. While NN attains high predictive accuracy, the 浙江万里学院现代物流学院 外文文献原文 3 incomprehensibility of its predictive behavior prevents its application in solving in inventory management problems. For this reason, this research is exploring the techniques and
21、methods to extract rules, embedded in the weights of a neural network which was trained to forecast demand and lead time in inventory management. 2. LITERATURE REVIEW This section provides a comprehensive study of three topics which are inventory management, knowledge discovery process and ANN. Inve
22、ntory Management requires detail planning, controlling and monitoring the flow of and storage of goods and service. To achieve satisfactory level of service, implicit and tacit knowledge is required so as to enhance inventory management. 2.1 Inventory Management is the art of managing inventory whic
23、h is defined as a physical resource that a firm/company holds in stock with the intent of selling it or transforming it into a more valuable state. Over the past fifteen years, inventory management has become an important focus of competitive advantage for firms and organizations. According to Sprag
24、ue, inventory management plays a crucial role to the enterprise because its performance improvement can enhance overall supply chain activities. Liberman identifies five inventory management performance factors and the challenge faced by the companies (1) Technical factors- Companies lack the availa
25、bility, reliability and knowledge of efficient technology; (2) Organizational factors- Companies may not have right technical input and financial support; (3) Financial factors- Companies may not have explicit financial mechanisms; (4) Managerial factors- Companies do not have sufficient training an
26、d proper management, and (5) Informational factors Companies lack appropriate information and have information sharing problems. Another study on factors influencing inventory management performance conducted by Vries has grouped inventory management performance as followings (1) Economic, (2) Behav
27、ioral and (3) Organizational categories. Similarly, in the research of Rajeev, he identifies factors such as: (1) Safety stock, (2) Capacity utilization level, (3) Purchasing effectiveness, (4) Demand forecasting accuracy, (5) Practices of inventory management, (6) Management attitude, (7) Employees
28、 training, (8) Interaction with suppliers, (9) Interaction with customers and, (10) Supplier empowerment. Other researchers narrowed the number of inventory management performance 浙江万里学院现代物流学院 外文文献原文 4 factors to technical aspect, and studied only the influential factors in forecasting. Safety Stock
29、 which is the extra units of inventory carried as protection against possible stock-outs. While making the order quotation and calculating production plan, the safety stock must be used to guard against the uncertainty of the manufacturing processes, demand and lead time. Some factors such as lead t
30、ime, number of customers, customer satisfaction, delivery reliability, supply reliability can affect the safety stock. Dongxu identified the factors such as use frequency, quality grade and actual safety stock, while Jia suggested forward factors such as stock-out cost, sales situation. Figure 1 is
31、an integration of all influential factors that determine safety stock. 3. METHODOLOGY This section proposes a knowledge discovery model shown in Figure 4 and a computational example using a given set of data . Table 1 is the extract of first 9 values from 48 total values. 浙江万里学院现代物流学院 外文文献原文 5 Knowl
32、edge Discovery Model used is composed from nine steps. This paper focuses only on first seven steps. Rule extraction step will be analyzed in depth in future work. Decompositional approach is based on an extraction algorithm in which each component of the network is examined. The knowledge extracted
33、 at this level is combined afterward. As this method concentrates on the individual components, this is considered an open-box approach. Pedagogical (input/output) approach concentrates on the analysis of the input /output behavior of the network. In essence, the method pertains to of the black box
34、approaches. The purpose of this experiment is to demonstrate that a neural network learns facts about relationships between the input variables and output ones, if the relation exists. Therefore after testing different NN architectures a NARX network is selected, with 17 neurons in the hidden layer
35、and having four variables as inputs: Period, Season factor, Level, Trend and Demand as output (Table 1). In the hypothesis, it is assumed that there is a relation between input variable and output 浙江万里学院现代物流学院 外文文献原文 6 safety stock (S) described by equation Winter equation : F t+l =(L t +lT t )S t+l
36、 , where F t+n is the forecasted demand for period l, L t is the level, T t is the trend, and L t+1=(D t+1/S t+1)+(1-)(L t +T t) T t+1= (L t+1 -L t)+(1-)T t S t+p+1 =(D t+1 /L t+1)+(1-)S t+1 To simplify the model we assume that T and S are constant, and the periodicity of demand is 4. For example th
37、e forecasted demand at period 2 is calculated as: F2 =(L1 +2T 1)S 2=(18439+2*524)0.68=13251.16 Therefore in training the network, the input layer is fed with 4 variables: L1, T1, l(time period) and S2, while the output is F2. The network uses the default Levenberg-Marquardt algorithm for training. T
38、he application randomly divides input vectors and target vectors into three sets: 60% are used for training, 20% are used to validate that the network is generalizing and to stop training before overfitting; the last 20% are used as a completely independent test of network generalization. 6. CONCLUS
39、ION The paper adopts computational example to show that a neural network is able to detect the hidden knowledge. Apart from detecting knowledge embedded in NN, NN can be used in discovering relations between inputs parameters successfully. However, knowledge extraction and representation in an under
40、standable way for humans needs further investigation. Therefore the further research is to explore how to represent knowledge related to inventory management and to investigate either Decompositional or Pedagogical approach is more appropriate in this study. Neural networks models shows promising re
41、sults in increasing the forecast accuracy, both in time series and causal models. Extracting the knowledge embedded in NN weights could help in understanding the relevance of input parameters and therefore the factors which are determining inventory management efficiency. The process employed with extracting knowledge from trained NN is Knowledge Discovery process. In this paper, the algorithm of knowledge discovery process has been analyzed by studying the steps responsible with collecting, processing and deploying of knowledge.
