Aprior算法的改进及其在高职学生就业中的应用

Apriori算法是一种最具有影响力的挖掘布尔关联规则频繁项集的算法。然而,在产生频繁项集时,需要多次扫描事务数据库,并产生庞大的候选集。为了降低算法的时间复杂度,提高程序运行效率,提出了一种改进的Apriori算法。同时,以某一高职院校中已毕业学生和在校生的就业信息为基础,建立了就业事务数据库,运用改进的Apriori算法对就业数据进行处理,产生了频繁项集。根据最小支持度和最小置信度的定义,找出强关联规则。并引入重要性对强关联规则进行分析,发现具有实际意义的信息,为高职学校就业指导提供决策,为学生就业提供了若干意见。     

面向计算机集群系统的FP-Growth算法的并行计算

FP-Growth是频繁模式挖掘的经典算法，能够在不产生候选集的情况下生成所有的频繁模式，效率与Apriori算法相比有巨大提高，然而FP—Growth算法在挖掘频繁模式过程中需要递归构建大量的条件FP—tree，并分别针对这些条件FP—tree进行挖掘，时间及空间效率不高，在实际应用中存在很大局限性。计算机集群是由多台普通计算机设备通过特定方式结合在一起构成的并行处理系统，属于分布式计算环境，具有计算能力强大、性价比高、灵活等优势。本文提出一种面向计算机集群的并行挖掘算法Gridify FP-Growth，该算法以FP—Growth为基础，通过任务划分的形式，将计算任务分配到计算机集群中各个计算节点上执行，充分利用各个节，最的计算资源，最后汇总各节点的计算结果。实验证明GridifyFP—Growth算法不会牺牲计算的准确性，并可以大幅度缩短计算时间，有效缓解计算大规模数据库时的内存压力。     

基于移动Agent的企业级数据挖掘应用模型研究

移动Agent可以看作是人工智能与分布式计算技术相结合的产物,已经成为分布式计算研究的热点。该文通过对移动Agent关键技术进行分析,探索了将移动Agent技术与分布式数据挖掘相结合的可行性,讨论了采用移动Agent技术进行分布式数据挖掘时所遇到的问题及解决方案,提出了一个基于IBM Aglets 2.0.2移动Agent系统平台的企业级分布式数据挖掘应用模型,并给出了数据挖掘agent算法。     

频繁采动影响下巷道修复技术的应用

介绍频繁采动影响下巷道和硐室,因支护方式和参数选择不尽合理,在施工修复和使用过程中时有破坏。采用近年来推广的锚、网、喷、梁、索、注支护技术进行修复巷道,在巷道、硐室的支护与修复开辟了一条新的途径。它的推广应用,有效地控制了围岩变形,保证了巷道、硐室支护的稳定。     

关联规则中支持度变化的数据重挖掘算法研究

使用关联规则进行数据挖掘时,使用者为了达到一定的挖掘效果,经常需要不断地改变关联规则的支持度阈值(support)。文中提出了一种从大型数据库中挖掘关联规则的快速算法。该算法以经典的Apriori算法为基础,可以在提出新的支持度后,直接在首次挖掘的基础上进行再一次挖掘。结果表明,它较Apriori算法的实现速度有明显的提高。     

A data mining algorithm for fuzzy transaction data

The main purpose of this paper is to propose a data mining algorithm for finding interesting association rules from given sets of fuzzy transaction data. To efficiently resolve the ambiguity frequently arising in available information and do more justice to the essential fuzziness in human judgment and preference, the trapezoidal fuzzy numbers are used to describe the fuzzy assessments of transaction data. Then, combining the concepts of fuzzy set theory and the priori algorithms, the interesting item sets are found to construct the association rules. Finally, a numerical example is used to demonstrate the computational process of proposed data mining algorithm. By utilizing this data mining algorithm, the decision-makers’ fuzzy assessments with various rating attitudes can be taken into account in the data mining process to assure more convincing and accurate knowledge discovery.     

在物流RFID数据库中挖掘时空模式

对传统的关联规则进行改进,结合时间和空间属性,提出了适合在物流RFID数据库中挖掘时空频繁模式的一些相关概念:频繁装(卸)货地点、频繁直达路径和频繁转运路径等。利用这些概念和有效的剪枝策略能快速准确地挖掘出物流RFID数据库中隐含的非平凡模式,可为物流公司的科学管理与线路优化提供及时准确的决策依据。     

A systematic mapping study of process mining

This study systematically assesses the process mining scenario from 2005 to 2014. The analysis of 705 papers evidenced ‘discovery’ (71%) as the main type of process mining addressed and ‘categorical prediction’ (25%) as the main mining task solved. The most applied traditional technique is the ‘graph structure-based’ ones (38%). Specifically concerning computational intelligence and machine learning techniques, we concluded that little relevance has been given to them. The most applied are ‘evolutionary computation’ (9%) and ‘decision tree’ (6%), respectively. Process mining challenges, such as balancing among robustness, simplicity, accuracy and generalization, could benefit from a larger use of such techniques.     

数据挖掘和数据仓库技术在DSS中的研究与应用

重点论述了数据挖掘及数据仓库技术,在此基础上分析了DM、DB、OLAP相结合的决策支持系统的架构与设计过程。对决策支持系统未来的应用方向作出了展望。     

A data-driven methodology for supporting resource planning of health services

In recent years, healthcare systems have been forced to better organize their services in the final attempt to maximize both care effectiveness and efficiency. In particular, emergent trends are prompting hospitals to pay more attention to the effective and efficient planning of resources and to the creation of patient-centred services, in which current activities and resources are reorganized around patients.This paper proposes a process mining based methodology to systematically support the resource planning of health services. Specifically, combining Time-Driven Activity Based Costing and process mining approaches, it automatically identifies the patient flow and analytically evaluates activities, service times, and resource consumptions for a specific class (-es) of patients (e.g., a DRG, patients with specific medical condition, etc.). Thus, it allows to reliably estimate the expected resource consumptions for the patient group under investigation. Thanks to process mining, the method overcomes the limitations of existing quantitative approaches that are often time-consuming, based on subjective observations, and too case specific. The method was applied to a real case study of lung cancer patients in an Italian hospital.     

基于分布式缓存机制与数据仓库的多式联运物流决策支持系统

为解决业务数据量的不断增加导致传统物流决策支持系统很难满足企业要求的问题,提出了基于分布式缓存机制与数据仓库的多式联运物流决策支持系统结构,给出了系统的工作流程,并且设计了高效缓存替换算法。该系统包括多数据源组成的基础层;模型库、知识库、数据仓库与式分布式缓存机制组成的处理层;联机分析处理和数据挖掘等组成的应用层,实现了定性与定量相结合的物流决策支持,提高了多式联运物流企业决策效率。     

Node localisation in wireless sensor network based on self-organising isometric embedding

An accurate distributed localisation method which uses the connectivity measurement to localise nodes in wireless sensor network is proposed in this paper. The proposed method is based on self-organising isometric embedding algorithm that adaptively emphasises the most accurate range measurements and naturally accounts for communication constraints within the sensor network. Each node adaptively chooses a neighbourhood of sensors to update its estimated position by minimising a local cost function and then passes this update to neighbour sensors. Simulation demonstrated that the proposed method is more robust to measurement error than previous proposals and it can achieve comparable results using fewer anchor nodes than previous methods. The applications of the new method in logistics and transportation system monitoring are also illustrated in the paper.     

一种基于Java及数据挖掘技术的学生管理系统

基于J2EE分布式体系结构的设计模式,针对大量的并发访问、复杂环境、网络不稳定等问题,提出了实现基于Java及数据挖掘技术的学生管理系统的解决方案。系统以Spring框架为核心,向下整合Hibernate框架进行持久层访问,向上整合Struts2框架按清晰的MVC(Model、View、Controller)模式控制,各层次之间以松耦合方式组织在一起,可清晰划分应用的层次,提高系统的灵活性,可扩展、可维护及可复用性。     

基于数据挖掘的模糊物流配送评价技术

提出了一种基于数据挖掘的物流配送系统模糊评价技术。以某类图书在湖南省的营销状况为例,系统全面地介绍了该类图书营销配送体系的评价过程。分析表明,该决策方案能有效地对客户资源进行分级管理,为企业ERP的建设提供了重要的决策依据。     

An adaptive distributed localisation in wireless sensor network

Many applications of wireless sensor networks (WSN) in industry can benefit from fine-grained localisation. In this paper, we propose an accurate, distributed localisation method which uses connectivity measurements to localise sensor nodes in WSN. The proposed method is based on a manifold learning embedding algorithm that adaptively emphasises the most accurate range measurements and naturally accounts for communication constraints within the WSN. Each node adaptively chooses a neighbourhood of sensor, updates its position estimate by minimising a local cost function and then passes this update to neighbouring sensors. Simulation results demonstrate that the proposed method is more robust to measurement errors than previous proposals and it can achieve comparable results using many fewer anchor nodes than previous methods.     

决策树算法在基本医疗保险中的应用研究

随着计算机技术以及学科交叉的发展,数据挖掘技术被越来越多应用到保险的风险分析中。文章使用数据挖掘基本算法之一的决策树算法分析研究了某市的基本医疗保险数据样本,找出隐含的有效信息,并在此基础上提出了完善基本医疗保险的对策建议。     

中观信息系统审计风险控制体系研究——以COBIT框架与数据挖掘技术相结合为视角

当前,我国急需一套完善的中观信息系统审计风险控制体系。这是因为我国的中观经济主体在控制信息系统审计风险时需要一套成熟的管理流程,且国家有关部门在制定信息系统审计风险防范标准方面也需要完善的控制体系作为支撑。在阐述COBIT与数据挖掘基本理论的基础上,借鉴COBIT框架,构建中观信息系统审计风险的明细控制框架,利用数据挖掘技术有针对性地探索每一个明细标准的数据挖掘路径,创建挖掘流程,建立适用于我国中观经济特色的信息系统审计风险控制体系。     

Recommendation system using feature extraction and pattern recognition in clinical care systems

Medical information systems have been increasingly facilitating and improving the quality of health monitoring, disease-trend modelling and early intervention with evidence-based medical treatment by data mining and feature extraction. Such systems are part of the enterprise information system of the healthcare organisations. We proposed a new algorithm fb-kNN towards recommendation algorithms based on analysis of the patterns of diseases with patterns in human body, which was then implemented in Healthcare 4.0 for the recommendation of diagnosis and treatment. Our developed tool is a complete package solution for the Enterprise Management System (ERP) which shows improvement in healthcare, reducing chronic diseases and mortality rates.     

Mining frequent itemsets: a perspective from operations research

Mining frequent itemsets is a flourishing research area. Many papers on this topic have been published and even some contests have been held. Most papers focus on speed and introduce ad hoc algorithms and data structures. In this paper we put most of the algorithms in one framework, using classical Operations Research paradigms such as backtracking, depth‐first and breadth‐first search and branch‐and‐bound. Moreover, we present experimental results where the different algorithms are implemented under similar designs.     

Flood decision support system on agent grid: method and implementation

In this paper, we introduce the concept and architecture of agent grid. Agent grid is an intelligent platform that enables the independent operating entities (agents) to interact with one another to form dynamic services on the Grid. Under this view, we built an agent grid platform named AGrIP that includes four layers and several useful toolkits. With the platform support, we implemented the flood decision support system which combines the wireless sensor network for data acquisition and software agent technology for legacy system integration. Additionally, we developed a toolkit for programmers to visually develop software agents which makes the development process easier. Besides, the MWAC model proposed is for sensor network to save power which can transit the information for long distance. This system is now applied as a module in the city emergency interact project.