中国管理研究与实践

出版时间：2011-12  出版社：复旦大学出版社  作者：石勇，唐立新，汪寿阳　著  页数：461  

内容概要

　　复旦管理学奖励基金会由复旦校友、原中共中央政治局常委、国务院副总理李岚清同志发起，成立于2005年9月。宗旨是奖励我国在管理学领域做出杰出贡献的工作者，倡导管理学理论符合中国国情，并密切与实践相结合，推动我国管理学长远发展，促进我国管理学人才的成长，提高我国管理学在国际上的学术地位和影响力。
　　复旦管理学奖励基金会设有“复旦管理学杰出贡献奖”，自2006年起，每年依次在管理学的三个子领域“管理科学与工程”、“工商管理”和“公共管理”进行评奖。2009年评奖领域是管理科学与工程，产生了三位获奖入：石勇、唐立新、汪寿阳。
　　《中国管理研究与实践：复旦管理学杰出贡献奖获奖者代表成果集（2009）》汇集了2009年复旦管理学杰出贡献奖的三位获奖者的代表性学术成果，这些成果代表着目前我国管理学研究的领先水平，在创新性、学术性、实用性三个方面达到了一流标准，对广大管理学研究者有很强的借鉴意义和理论价值。
　　本书适用于高等院校管理学领域的研究者，也可作为政府经济管理部门工作人员、从事企业管理工作的基层管理者的参考用书。

作者简介

石勇，教授是管理科学与工程领域的国际知名学者。在模糊数学理论、多目标多资源数学规划理论、系统框架和应用、智能知识管理等领域作出了突出的贡献。主要体现在以下方面。(1)创立了以多目标线性规划为基础的信用评分系统与数据挖掘理论和方法用多目标最优化概念解决数据挖掘中不同数据同时分割的标准问题。该方法已被广泛应用在信用评分管理、金融信贷风险控制、信息侵扰和侦测、水资源管理、生物信息学、石油勘探及采掘等领域。(2)使用最优化理论对“数据－数据挖掘－智能知识管理”这一崭新的交叉性国际科学领域进行研究，首先提出了“智能知识管理”的概念，把数据挖掘的结果作为智能知识的基础，探讨了智能知识与传统知识结构之间的逻辑关系，研究了数据挖掘与智能知识内在联系的数学模型并用其展示智能知识的特征。智能知识研究不仅在数据挖掘与知识管理的交叉学科间推动并促进了传统管理科学的发展，而且具有重要的实际管理应用背景。石勇教授与他的科研团队先后将他们的理论成果应用在中国人民银行个人征信评分系统、网易公司客户管理、中国工商银行的优质客户管理、中国再保险公司的战略管理及中国投资公司的投资匹配决策等影响中国经济发展的重大项目中，其成果所带来的经济价值和社会价值正在不断显现。(3)创立多目标多资源线性规划理论、方法及应用的数学及理论框架，该成果以多目标多资源线性规划为基础，广泛研究了多资源水平对已有的最优线性系统设计的影响。提出了线性系统设计不应是寻求给定系统的一个最优点，而应基于资源水平设计一个最优系统。这些研究成果已在会计价格转移、资金预算、生产计划、管理信息系统和通讯管理等方面得到了成功的实际应用。

书籍目录

一、　石勇学术成果汇集篇
个人简介
Several Multi?criteria Programming Methods for Classification
Multiple Criteria Mathematical Programming for Multi?class
Classification and Application in Network Intrusion Detection
A Multi?criteria Convex Quadratic Programming Model for Credit Data
Analysis
Classifying Credit Card Accounts for Business Intelligence and
Decision Making：a Multiple?criteria Quadratic Programming
Approach
Data Mining Via Multiple Criteria Linear Programming：　Applications
in Credit Card Portfolio Management
Optimal System Design with Multiple Decision Makers and Possible
Debt：A Multicriteria De Novo Programming Approach
二、　唐立新学术成果汇集篇
个人简介
The Charge Batching Planning Problem in Steelmaking Process Using
Lagrangian Relaxation Algorithm
Simultaneously Scheduling Multiple Turns for Steel Color?coating
Production
Decision Support System for the Batching Problems of Steelmaking
and Continuous?casting
Production
Scheduling A Single Semi?continuous Batching Machine A Mathematical
Programming Model and Solution For Scheduling Production Orders in
Shanghai Baoshan Iron and Steel Complex
A Review of Planning and Scheduling Systems and Methods for
Integrated Steel Production
A Mathematical Programming Model for Scheduling
Steelmaking?continuous Casting Production
三、　汪寿阳学术成果汇集篇
个人简介
Why Does Energy Intensity Fluctuate in China？
Evolving Least Squares Support Vector Machines for Stock Market
Trend Mining
An Empirical Study on Information Spillover Effects Between the
Chinese Copper Futures Market and Spot Market
Currency Crisis Forecasting with General Regression Neural
Networks
An Integrated Data Preparation Scheme for Neural Network Data
Analysis
Crude Oil Price Forecasting with TEI@I
Methodology
A Novel Nonlinear Ensemble Forecasting

章节摘录

版权页：插图：As the development of information technologies and extensive use of internet, E-mail,one of basic functions provided by internet, has been a necessary tool for communication,so far. Because of its fastness, convenience and online function, a great number of firms and individuals have used E-mail to do some very important things, e.g. business deal,exchange private, secret and important information, etc. These firms and individual sexpect more safe, more reliable and more individualized E-mail service. Free mail-boxprovided by many websites cannot satisfy this demand. Because of this, many websites,for example, Sina, Netease, Yahoo etc. , provided VIP E-mail box, which is a chargedservice. It was studied by Chineselabs and iResearch [23, 24] that the total market size ofChinese VIP E-mail service was up to 5. 1 hundred million RMB in 2004 and will reach 6.4 hundred million RMB in the future.  The key problem faced by the VIP-mail suppliers is toidentify the easy-to-leave customers and to exploit some methods to attract them to usetheir VIP E-mail. In this study, the customers are classified into two groups: lost （meanseasy-to-leave） and current （means permanent user）. The procedure is as in Section 4.2. The difference is that the training dataset is selected with 500 Lost records and 500 Currentrecords. The VIP E-mail dataset used in this paper is provided by a famous website hosting company in China.  Original customer database is mainly composed of automated machine recorded customer activity journal and large amount of manually recorded tables. These data are distributed among servers located in different departments of this company, covingmore than 30 kinds of transaction data charts and journal document9 with over 600 attributes. If we were to directly analyze these data, it would lead to “curse ofdimensionality", that is, the drastically raising of computational complexity and classification error due to the increase of dimension of data.  Hence, the dimensionality of the feature space must be reduced before classification. Feature extraction （FE） is one of the dimensionality reduction techniques. FE extracts a subset of new features from theoriginal feature set by means of some functional mapping.

编辑推荐

《中国管理研究与实践:复旦管理学杰出贡献奖获奖者代表成果集(2009)》由复旦大学出版社出版。