改进ID3算法在管道腐蚀检测系统中的应用研究

ID3算法计算速度较快、容易实现并且适用于处理规模较大的学习问题,但其较倾向于选择取值较多的类别,从而导致丢失全局最优解。提出一种改进的ID3算法,并将其应用于管道腐蚀检测系统中,研究结果表明,改进后的算法具有较高的可信度。     

Caller ID and the meaning of privacy

Caller ID service continues to be a controversial issue in the U.S. because of its privacy implications. State and federal regulators, legislators, scholars, and the courts have examined and responded to the privacy issue from a policy perspective, but perhaps without a complete understanding of the meaning of privacy in the context of the debate. What types of privacy are involved, how significant are these interests, and how might privacy needs compare and be balanced? This article explores privacy in the context of the Caller ID debate from a social science perspective. It examines motives for seeking and preserving privacy and explores the dynamic relationship between the caller and called party positions. It then provides an analysis of current and proposed Caller ID features and policies with a view toward understanding how these proposals balance competing privacy needs. This article establishes an analytic framework and a foundation for further study of caller and called party privacy that should lead to a better understanding of the privacy debate and the privacy implications of Caller ID.     

数据挖掘中决策树分类算法的研究与改进

决策树分类算法是数据挖掘中一个重要的内容,而ID3算法又是决策树分类算法中的一种重要方法且被广泛应用。然而在实际应用过程中,现存的决策树算法也存在着很多不足之处,如计算效率低下、多值偏向等。为了解决这些问题,提出了一种基于ID3算法的加权简化信息熵算法,它提高了决策树的构建速度,减少了算法的计算运行时间,同时也克服了ID3算法往往偏向于选择取值较多的属性作为测试属性的缺陷。并且随着数据规模的增大,决策树的分类性能表现得越好。     

基于ID3算法的企业产品安全库存量预测

通过构建企业数据仓库,对数据仓库中的数据进行归约预处理后, 得到训练样本集, 采用ID3 算法建立一棵企业客户购买力分类决策树, 通过对企业客户购买力的分析,科学地预测企业产品的安全库存量.     

Surveillant Institutional Eyes in Korea: From Discipline to a Digital Grid of Control

This article investigates the shift from disciplinary societies (the visible and physical violence of power) to control societies (the modulating and normalizing techniques of power) in South Korea. At the institutional level, during the period of repressive and disciplinary society in Korea (1948-1992), the regulatory control systems of the state were mainly performed by two formidable apparatuses: the national ID system and the National Security Law. On the other hand, the deployment of institutional power since 1993 has been based on the logic of free-floating control, dispersion, normalization, and modulation. This study examines how the techniques of power were gradually transformed from a centralized and hierarchical model into a distributive and dispersed network model, based on flow, speed, and mobility.     

竞技健美操运动员踝关节力量要求及训练方法的研究

采用文献资料、专家访谈、问卷调查、逻辑分析等研究方法,对踝关节解剖结构、竞技健美操运动员踝关节损伤现状、竞技健美操运动员踝关节力量要求及训练方法进行了调查、分析、归纳与总结。研究结果表明：人体踝关节解剖结构复杂,易出现该部位的运动损伤,竞技健美操运动项目特征对踝关节力量素质提出了较高的要求,应用赤脚加速跑、节奏性纵跳等练习提高学生的踝关节力量素质,进而达到提高竞技健美操运动员技术水平的目的。     

基于熵的决策可信度分析

从分类的角度进一步阐述了信息熵的概念，使得利用信息熵的方法构造决策树有了更加坚实的数学基础，研究了信息熵与决策可信度的关系，给出了利用信息熵计算决策最大和最小可信度简单、直观易操作的算法，在期望信息熵较小时计算得出的最大和最小可信度是非常接近的，因此可以用它来做估计决策可信度的依据，实验结果也表明计算的结果和实际的决策可信度是非常接近的。     

集资诈骗罪与非法吸收公众存款罪比较研究

近些年来,集资诈骗和非法吸收公众存款等非法集资活动多有发生。此类活动,一旦投资人的资金无法获得偿还,将会引发严重的社会不稳定。同时,大量社会资金循环于金融机构之外,将会严重干扰我国的金融秩序,不利于经济的稳定。对该类违法犯罪行为既要严惩,更要采取有效措施加大防范力度。     

分类决策树在交通事故成因分析中的应用

交通事故的形成常涉及多种因素,如驾驶人本身、车辆、环境等.以数据挖掘技术中的ID3分类决策树算法对交通事故数据进行分析和研究,得出蕴含在数据中的有益模式,决策树构造结果显示,分类决策树应用于交通事故成因分析中具有可行性与有效性.