支持向量机在天然气企业产量预测中的应用

天然气企业副产品产量受市场的波动影响很大,对于副产品产量的预测,通常都是依靠工作人员的经验,工作量大,效率不高。回归统计分析方法及传统的机器学习方法在这个领域的应用存在着一些局限性,介绍一种新的机器学习算法—支持向量机,以企业为背景,运用支持向量机算法来解决预测问题,实验证明采用支持向量机能够满足企业的需求,得到满意的效果。     

最小二乘支持向量机在睡眠打鼾诊断中的应用

支持向量机是数据挖掘和机器学习领域中的重要方法之一,最小二乘支持向量机是支持向量机学习算法的重要扩展,在训练速度方面有明显优势。对支持向量机现有的多类分类算法（一对一方法、一对多方法、纠错输出编码方法和最小输出编码方法）引入了最小二乘支持向量机,并应用于睡眠打鼾疾病的诊断预测中,取得了较好的效果。     

应用于数据挖掘分类算法的SVM研究

本文主要介绍了支持向量机的基本思想，通过目前SVM训练算法的研究成果分析了它在数据挖掘中（尤其是分类算法上）的应用，并阐述了支持向量机在数据挖掘领域中实现的方法。     

一种有效的供应链绩效预测方法及其应用

从知识发现和数据挖掘的角度,利用粗糙集和支持向量回归机的理论和方法,建立了基于粗糙集和支持向量回归机相结合的供应链绩效预测模型。结合一个供应链绩效预测实例,首先对其基于平衡记分卡的指标体系进行了约简,然后将约简的评价指标输入到支持向量回归机中进行训练,构建预测模型,最后把预测的样本输入到模型中进行供应链绩效预测,预测结果与实际结果基本吻合。     

支持向量机在商场客流预测中的应用

支持向量机是数据挖掘中的一个新兴技术,它对非线性的决策边界间的建模能力是高度准确的.本文通过分析支持向量机的原理和算法,并给出基于支持向量机的客流量预测模型,最后通过试验结果说明了支持向量机在预测中的有效性.     

基于支持向量机的分类预测算法研究

分类预测是数据挖掘、机器学习和模式识别等很多领域共同关注的问题,已经存在了许多有效的分类算法,但这些算法还不能解决所有的问题。支持向量机作为一种新的分类预测工具,能根据有限样本信息在模型的复杂性和学习能力间取得平衡,并能获得更好的泛化能力。SMO算法是支持向量机中使用最多的算法,它体现了支持向量机的优点,同时也能处理大规模训练集。     

基于支持向量机的卷烟焦油预测

分析了卷烟焦油含量预测问题,提出了基于支持向量机的卷烟焦油含量预测方法。首先,介绍了支持向量回归估计的学习算法。其次,建立了基于支持向量机的卷烟焦油含量预测模型。然后,提出了卷烟焦油含量支持向量机预测的实现算法。最后给出了一个算例。实例表明,该方法能够根据烟叶中的化学成分的测量值来预测卷烟的焦油含量。     

多类支持向量机文本分类方法

文本分类是数据挖掘的基础和核心,支持向量机（SVM）是解决文本分类问题的最好算法之一。传统的支持向量机是两类分类问题,如何有效地将其推广到多类分类问题仍是一项有待研究的课题。介绍了支持向量机的基本原理,对现有主要的多类支持向量机文本分类算法进行了讨论和比较。提出了多类支持向量机文本分类中存在的问题和今后的发展。     

多类支持向量机文本分类方法

文本分类是数据挖掘的基础和核心,支持向量机(SVM)是解决文本分类问题的最好算法之一.传统的支持向量机是两类分类问题,如何有效地将其推广到多类分类问题仍是一项有待研究的课题.介绍了支持向量机的基本原理,对现有主要的多类支持向量机文本分类算法进行了讨论和比较.提出了多类支持向量机文本分类中存在的问题和今后的发展.     

基于核函数的支持向量机分类方法

支持向量机是目前正在兴起的一种新的数据挖掘分类方法,阐述了支持向量机的理论基础及核函数,阐明了支持向量机分类的基本思想,分析了支持向量机的优缺点,对支持向量机在海量数据分类中的应用前景进行了展望。     

需求损失下两产品联合生产动态批量决策及预测时阈

研究需求损失下两产品联合生产(采购)动态批量决策问题.在各周期成本变动情形下分析多周期动态批量决策的预测时阈和决策时阈,构建包含联合启动成本、两产品的单独启动成本、库存持有成本、变动生产成本和需求损失成本在内的成本最小化模型.在最优解结构特性的基础上,设计出前向动态规划算法求解问题,通过建立两产品生产点的单调性和建立生产集,给出求解预测时阈和决策时阈的充分条件.通过数值算例分析预测时阈求解的具体过程,表明所构建模型的有效性.     

Inventory Minimization and Production Scheduling: the Case of a Large Hong Kong Company

The background of this study is a rather classical but complex inventory control/production planning/line scheduling problem of a major soft-drink company in Hong Kong. The issue that stands out for this many-product high-sales manufacturer is the storage space of its central warehouse, which often finds itself in the state of overflow or near capacity with finished goods and work-in-process inventory. This phenomenon can create immediate interruptions of production, capital tie-ups and subsequent potential of lost sales. Another obviously important concern is the meeting of forecast demands. A mathematical modelling approach that entails techniques of multi-period aggregate optimization is proposed to tackle the overall problem. The dual objectives are to achieve better production planning and line scheduling in order to minimize inventory build-up and maximize demand satisfaction. Numerical results for a sample problem are reported as an illustration to this proposed two-phase approach.     

Measuring decision flexibility in production planning

We present a measure of decision flexibility for production planning problems. The flexibility of a decision is related to the size of the choice set associated with this decision. In production planning problems this set is a convex polyhedron in an n-dimensional space. Our measure of flexibility is the volume of this set with additional information about the shape. By using a new method recently proposed we give an analytical expression of the measure of flexibility. Sensitivity of flexibility to various parameters is also given analytically as a byproduct. Experimental results on a simple inventory problem are provided. An alternative measure of flexibility for large-scale syslems is also presented and discussed.     

用计划物料改善代加工企业的MRP

有很多制造型企业既存在自己制造生产产品,又存在为其他外部企业提供代为加工的生产模式。对于这样的企业,既是企业上了ERP系统,但是由于很难预估将为外部企业的代加工量,所以对于库存的原材料贮备量就难以估计,相应的客户成品的生产计划也难以制定。为了解决这一问题,防止对于加工原料及客户加工成品的储备量过高,提高产品的生产计划准确度,本文介绍一种采用计划物料的计划策略,它可以结合原有的生产物料清单(BOM)解决这一问题。     

Dual constrained single machine sequencing to minimize total weighted completion time

We study a single-machine sequencing problem with both release dates and deadlines to minimize the total weighted completion time. We propose a branch-and-bound algorithm for this problem. The algorithm exploits an effective lower bound and a dynamic programming dominance technique. As a byproduct of the lower bound, we have developed a new algorithm for the generalized isotonic regression problem; the algorithm can also be used as an O(nlogn)-time timetabling routine in earliness-tardiness scheduling. Extensive computational experiments indicate that the proposed branch-and-bound algorithm competes favorably with a dynamic programming procedure. Note to Practitioners-Real-life production systems usually involve multiple machines and resources. The configurations of such systems may be complex and subject to change over time. Therefore, model-based solution approaches, which aim to solve scheduling problems for specific configurations, will inevitably run into difficulties. By contrast, decomposition methods are much more expressive and extensible. The single-machine problem and its solution procedure studied in this paper will prove useful to a decomposition method that decomposes multiple-machine, multiple-resource scheduling problems into a number of single-machine problems. The total weighted completion time objective is relevant to production environments where inventory levels and manufacturing cycle times are key concerns. Future research can be pursued along two directions. First, it seems to be necessary to further generalize the problem to consider also negative job weights. Second, the solution procedure developed here is ready to be incorporated into a machine-oriented decomposition method such as the shifting bottleneck procedure.     

Handling qualitative aspects in Unequal Area Facility Layout Problem: An Interactive Genetic Algorithm

The Unequal Area Facility Layout Problem (UA-FLP) has been addressed using several methods. However, the UA-FLP has only been solved for criteria that can be quantified. Our approach includes subjective features in the UA-FLP, which are difficult to take into account with a more classical heuristic optimization. In this respect, we propose an Interactive Genetic Algorithm (IGA) that allows an interaction between the algorithm and the Decision Maker (DM). Involving the DM's knowledge in the approach guides the search process, adjusting it to his/her preferences at each generation of the algorithm. In this paper, we are concerned with assisting the DM in finding a good solution according with criteria that can be: subjective, unknown at the beginning or changed during the process, so that, the problem addressed differs from a classic optimization problem. In order to avoid overloading the DM, the whole population is classified into clusters by the fuzzy c-means clustering algorithm and only one representative element of each cluster is directly evaluated by the DM. A memory of the best solutions chosen by the DM is kept as a reference. The tests carried out show that the proposed IGA is capable of capturing DM preferences.     

基于人工神经网络的中长期负荷预测算法

当前中长负荷预测的大部分方法都衍生于传统的线形统计理论,难以解决复杂的非线性问题。文中结合BP人工神经网络技术,利用人工神经网络所具有的非线性映射和函数逼近功能对中长期电力负荷进行了研究,提出了一种中长期电力负荷预测的思路。并利用北京市的实际数字对未来若干年的用电量进行了预测,实验结果表明,该算法具有较好的准确性和可行性。     

Some classes of preference choice rules for decision-making problems

The topic under discussion is modeling the subject of a decision making system, a decision maker (DM), up to the information, which it relies upon while making a concrete decision, directing towards the purpose before it, i.e., choosing the best possible action. Clearly, the indicated model of a DM substantially depends on the object of decision making, the situation of decision making (SDM), which this DM roughly presents in the form of the so-called situation of decision problem (SDP) through discarding the decisions impossible (or not interesting) for itself as well as impossible (according to its view) consequences from the initial situation, thus obtaining the so-called scheme of the situation of decision problem (SSDP). The interchangeability of the obtained models is studied.     

A heterogeneous computing system for data mining workflows in multi-agent environments

Abstract: The computing-intensive data mining (DM) process calls for the support of a heterogeneous computing system, which consists of multiple computers with different configurations connected by a high-speed large-area network for increased computational power and resources. The DM process can be described as a multi-phase pipeline process, and in each phase there could be many optional methods. This makes the workflow for DM very complex and it can be modeled only by a directed acyclic graph (DAG). A heterogeneous computing system needs an effective and efficient scheduling framework, which orchestrates all the computing hardware to perform multiple competitive DM workflows. Motivated by the need for a practical solution of the scheduling problem for the DM workflow, this paper proposes a dynamic DAG scheduling algorithm according to the characteristics of an execution time estimation model for DM jobs. Based on an approximate estimation of job execution time, this algorithm first maps DM jobs to machines in a decentralized and diligent (defined in this paper) manner. Then the performance of this initial mapping can be improved through job migrations when necessary. The scheduling heuristic used considers the factors of both the minimal completion time criterion and the critical path in a DAG. We implement this system in an established multi-agent system environment, in which the reuse of existing DM algorithms is achieved by encapsulating them into agents. The system evaluation and its usage in oil well logging analysis are also discussed.