含非连续性信息多属性案例中的决策知识发现方法

中国管理科学 ›› 2014, Vol. 22 ›› Issue (4): 83-91.

含非连续性信息多属性案例中的决策知识发现方法

梁昌勇¹, 顾东晓^1,3, 程文娟¹, 杨昌辉^1,4, 顾佐佐²

1. 合肥工业大学管理学院, 安徽合肥 230009;
2. 安徽大学艺术与传媒学院, 安徽合肥 230011;
3. 南京大学信息管理学院, 江苏南京 210093;
4. 安徽省产业转移与创新发展人文社会科学重点研究基地, 安徽合肥 230009

收稿日期:2012-06-30 修回日期:2013-05-07 出版日期:2014-04-20 发布日期:2014-04-23
作者简介:梁昌勇（1965- ），男（汉族），安徽肥西人，合肥工业大学管理学院教授，博士生导师，研究方向：管理信息系统、智能决策方法、行为决策等.
基金资助:
国家自然科学基金资助项目（71331002，71301040，71271072，71171072，51274078）；中国博士后科学基金面上项目（2013M541651）；安徽省社会科学重点研究基地重点项目（SK2013A148，2013AJRW0131）；合肥工业大学青年教师创新项目（2013HGQC0026）

A Decision Knowledge Discovery Method for Multi-attribute Cases with Non-continuous Features

LIANG Chang-yong¹, GU Dong-xiao^1,3, CHENG Wen-juan¹, YANG Chang-hui^1,4, GU Zuo-zuo²

1. School of Management at Hefei University of Technology, Hefei 230009, Chian;
2. School of Arts and Media, Anhui University, Hefei 230011, China;
3. School of Information Management at Nanjing University, Nanjing 210093, China;
4. Anhui Provincial Industrial Transfer and Innovation Development Key Research Institute of Humanities Social Science, Hefei 230009, China

Received:2012-06-30 Revised:2013-05-07 Online:2014-04-20 Published:2014-04-23

摘要/Abstract

摘要： 医疗决策案例中非连续性属性信息大量存在，含该类信息的案例知识发现是多属性案例决策的关键和难点。该文研究了含非连续性属性信息案例中的决策知识发现，将条件概率和GAs融合技术整合到案例推理方法之中，开发了KNN的延伸方法——CRMGACP法。该方法的核心是基于Gas进行权重获取和基于融合条件概率的改进相似度算法进行案例知识获取。在某大型综合医院收集数据，获取有效数据300条，基于VC++开发实现的BC-CBRsys平台进行了实验研究，结果表明CRMGACP比其他常见方法具有更好的性能，在多个统计指标上展示出显著的优势。显然，改进的案例决策方法克服了含非连续性信息案例决策知识难以获取的问题，在临床决策领域具有广阔的前景。

关键词: 复杂多属性决策, 公共卫生管理, 知识发现, 离散变量, 案例推理

Abstract: The information with non-continuous features is ubiquitous in diagnosis and treatment decision making cases. The knowledge acquisition of the cases with this kind of feature has always been a key and bottleneck in multi-attribute case decision making. In this paper, conditional probability and GAs are integrated into case-based reasoning technology to develop an extension method of traditional KNN——CRMGACP algorithm, which includes a GAs-based weight determination method and an improved similarity algorithm integrating the conditional probability. Collecting data from AH Hospital, which is one of large-scale hospitals in Anhui province, CancerCBRSys is developed as the experimental tool for tests. Experimental study is competed by comparing the performance amongst four different case-based reasoning methods. The results show that CRMGACP has the best performance and shows significant advantage in various statistics. In general, CRM-GACP solves the problem of knowledge discovery from non-continuous cases and is hopeful to be a powerful decision-making tool in the research area of clinical decision making.

Key words: complex multi-attribute decision making, public healch mangement, knowledge discovery, discrete variable, case-based reasoning

中图分类号:

C931.6

梁昌勇, 顾东晓, 程文娟, 杨昌辉, 顾佐佐. 含非连续性信息多属性案例中的决策知识发现方法[J]. 中国管理科学, 2014, 22(4): 83-91.

LIANG Chang-yong, GU Dong-xiao, CHENG Wen-juan, YANG Chang-hui, GU Zuo-zuo. A Decision Knowledge Discovery Method for Multi-attribute Cases with Non-continuous Features[J]. Chinese Journal of Management Science, 2014, 22(4): 83-91.

参考文献

[1] Ayer T, Chhatwal J, Alagoz O, et al. Informatics in radiology: Comparison of logistic regression and artificial neural network models in breast cancer risk estimation[J]. Radio Graphics, 2010, 30(1): 13-22.

[2] Hung M S, Shanker M, Hu M Y. Estimating breast cancer risks using neural networks[J]. Journal of the Operational Research Society, 2002, 53(2):222-231.

[3] Lieber J, Bresson B. Case-based reasoning for breast cancer treatment decision helping[C]. Blanzieri E, Portinale L. Advances in Case-Based Reasoning-Proceedings of the fifth European Workshop on Case-Based Reasoning (EWCBR-2k), Berlin: Springer, 2000.

[4] Floyd Jr C E, Lo J Y, Tourassi G D. Case-based reasoning computer algorithm that uses mammographic findings for breast biopsy decisions[J]. American Iournal of Roentgenol, 2000, 175(5):1347-1352

[5] Bilska A O, Floyd Jr C E. Investigating different similarity measures for a case-based reasoning classifier to predict breast cancer[J]. Imaging, 2001, 4322:1862-1866.

[3] 赵卫东, 盛昭瀚. 基于案例推理的决策问题求解研究[J].管理科学学报, 2000, 3(4): 29-36.

[4] 梁昌勇, 顾东晓. 面向不确定多属性决策问题的范例检索算法研究[J]. 中国管理科学, 2008, 17(1):131-137.

[5] 路云, 吴应宇, 达庆利.基于案例推理技术的企业经营决策支持模型设计术[J].中国管理科学, 2005, 13(2): 81-87.

[6] Gu Dongxiao, Liang Changyong, Li Xingguo, et al. Intelligent technique for knowledge reuse of dental medical records based on case-based reasoning[J]. Journal of medical systems, 2010, 34(2): 213-222.

[7] McCane B, Albert M. Distance functions for categorical and mixed variables[J]. Pattern Recognition Letters, 2008, 29(7): 986-993.

[8] Stanfill C, Waltz D L. Toward memory-based reasoning[J]. Communication of the ACM, 1986, 29(12): 1213-1228.

[9] Wilson D, Martinez T. Improved heterogeneous distance functions[J]. Journel of Artificial Intelligence Research, 1997, 6:1-34.

[10] Gower J C. A general coefficient of similarity and some of its properties[J]. Biometrics, 1971, 27: 857-874.

[11] Greene D, Freyne J, Smyth B. Padraig cunningham: An analysis of research themes in the CBR conference literature. ECCBR, 2008, 5239: 18-43.

[12] Renauda J, Levratb E, Fonteixc C. Weights determination of OWA operators by parametric identification[J]. Mathematics and Computers in Simulation, 2008, 77: 499-511

[13] Zhang Lu, Coenen F, Leng P. Formalising optimal feature weight setting in case-based diagnosis as linear programming problems[J]. Knowledge-Based Systems, 2002, 15(7): 391-398.

[14] Park C S, Han I. A case-based reasoning with the feature weights derived by analytic hierarchy process for bankruptcy prediction[J]. Expert Systems with Applications, 2002, 23 (3): 255-264.

[15] Dogan S Z, Arditi D, Gnaydin H M. Using decision trees for determining attribute weights in a case-based model of early cost prediction[J]. Journal of Construction Engineering and Management, 2008, 134 (2): 146-152.

[16] Fidelis M V, Lopes H S, Freitas A A. Discovering comprehensible classification rules with a genetic algorithm[C]. Proceeding of the 2000 Congress on Evolutionary Computation, La Jolla, July 16-19, 2000.

[17] Goldberg D E. Genetic algorithms in search, optimization and machine learning[J]. Machine Learning, 1988, 3(2-3):95-99.

[18] Cunningham P. A taxonomy of similarity mechanisms for case-based reasoning[J]. IEEE Transactions on Knowledge and Data Engineering, 2008, 21(11):1532-1543.

[19] West D, Mangiameli P, Rampal R, et al. Ensemble strategies for a medical diagnosis previous termdecisionnext term support system: A breast previous term cancer next term diagnosis application[J]. Europeon Journal of Operatinal Research, 2005, 162(2): 532-551.

[20] Lee S. Using data envelopment analysis and decision trees for efficiency analysis and recommendation of B2C controls[J]. Decision Support Systems, 2010, 49(4): 486-497.

[21] Salchenberger L M, Cinar E M, Lash N A. Neural networks: A new tool for predicting thrift failures[J]. Decision Sciences, 1992, 23(4): 899-916.

[22] Karabatak M, Cevdet Ince M. An expert system for detection of breast cancer based on association rules and neural network[J]. Expert Systems with Applications, 2009, 36(2): 3465-3469.

[23] Sexton R S, Dorsey R E. Reliable classification using neural networks: A genetic algorithm and back propagation comparison[J]. Decision Support Systems, 2000, 30(1): 11-22.

[24] Zhang Zhen, Zhang Hong, Bast Jr R C. An application of artificial neural networks in ovarian cancer early detection[C]. Proceedings of the IEEE-INNS-ENNS International Joint Conference on Neural Networks, Como, July 24-27, 2000.

[25] Tan T Z, Quek C, See Ng G, et al. Ovarian cancer diagnosis with complementary learning fuzzy neural network[J]. Artificial Intelligence in Medicine, 2008, 43(3): 207-222.

[26] Chen Yuehui, Wang Yan, Yang Bo. Evolving hierarchical RBF neural networks for breast cancer detection[J]. ICONIP, 2006, (3):137-144.

[27] Cruz-Ramirez N, Acosta-Mesa H G, Carrillo-Calvet H, et al. Discovering interobserver variability in the cytodiagnosis of breast cancer using decision trees and Bayesian networks[J]. Applied Soft Computing, 2009, 9(4): 1331-1342.

[28] Antal P, Verrelst H, Timmerman D, et al. Bayesian networks in ovarian cancer diagnosis: Potentials and limitations[C]. Proceedings of the 13th IEEE Symposium on Computer-Based Medical Systems (CBMS'00), Houston, June 23-24, 2000.

[29] Chang C L, Hsu Mingyuan. The study that applies artificial intelligence and logistic regression for assistance in differential diagnostic of pancreatic cancer[J]. Expert Systems with Applications, 2009, 36(7): 10663-10672.

[30] Witten I H, Frank E. Data mining: Practical machine learning tools and techniques[M]. Morgan Kaufmann: Elsevier, 2005.

[31] Venkatesh V, Speier C, Morris M G. User acceptance enablers in individual decision making about technology: Toward an integrated model[J]. Decision Sciences, 2002, 33(2): 297-316.

[32] Wixom B H, Todd P A. A theoretical integration of user satisfaction and technology acceptance[J]. Information Systems Research, 2005, 16(1): 85-102.

[33] Venkatesh V, Morris M G. Why don't men ever stop to ask for directions? Gender, social influence, and their role in technology acceptance and usage behavior[J]. MIS Quarterly, 2000, 24(1): 115-139.

[34] 陈浪涛, 张成洪, 张诚. 协同商务环境下基于案例推理机制研究[J]. 复旦学报(自然科学版), 2005, 44(6): 1009-1015.