中医辩证量化方法学研究
|
摘要
基于从定性描述到定量分析的科学发展一般规律，现代医学科学遵循着此规律取得了长足进步和快速发展。作为生命科学重要组成部分的中医学，纳入量化分析方法与手段，也已成为中医现代化的必然趋势。其中，中医辨证的量化研究，是当前提高中医药临床评价水平的关键科学问题。为此，引入最新发展起来的复杂事件数学分析模型贝叶斯网络，遵循中医学基本原理，在保持中医整体观特色前提下，就中医辨证量化方法进行了系统的研究。
首先，就中医辨证量化方法学的研究现状进行了系统分析。从辨证思维方法开始，总结了辨证的整体性、发散性、直觉性、形象性思维方法在中医辨证过程中的应用情况，继而分析了迄今为止不同作者在中医辨证量化研究中所用方法的优缺点及其效果。作为辨证量化的数理基础，模糊性判断、半定量方法，多因素分析、人工智能技术均已得到了大量的应用。但是，由于各自的缺陷，无一能理想地运用于中医辨证的量化研究。随着人工智能技术的逐步完善和推广，其在本研究领域的应用已成为新的发展趋势。
其次，对中医辨证思维规律与方法进行了探讨，从分析几种主要辨证方法之间的关系入手，根据导师朱文锋教授所创辨证统一体系理论，确定了辨证思维的关键环节——辨证要素，即病位与病性，讨
论了辨证要素对于建立辨证统一体系的作用及其在辨证定量研究中
的应用前景。因此；在症状与证候名称规范的基础上，探求症状／体
征对辨证要素的贡献度，探讨症状一要素一证侯辨证统一体系的完善
与实施方案，尤其是进一步将人工智能技术与之相结合，以应用于辨
证量化研究。
基于以上分析和人工智能数理模型的新发展，考虑到贝叶斯网
络原理与人脑思维模式，特别是辨证思维过程的良好拟合性，较深入
探讨了该模型应用于中医辨证量化研究的可行性。从网络结构、概率
分析到网络的自学习与经验积累，学习算法及流程图的构建，整个过
程与中医辨证思维与推理具有良好的吻合性。为4b，设计构造了基于
贝叶斯网络的中医量化辨证系统。
为考证该系统的实用效果，本文以一组806例肺系疾病住院病
例资料为样本，分别考察不同样本数贝叶斯网络自学习和预测效果。
结果表明，当样本数达到叩0时，量化辨证的预测就已达到较佳效果，
与人脑辨证思维过程拟合度较好。同时，以此研究样本为基础，分析
了过程中主要相关模块的工作原理与方法。
本研究结果表明，不仅中医辨证量化研究可以收到很好的效果，中医
学体系的量化发展也是可能的。
It is a natural tendency for Traditional Chinese medicine (TCM), as an very important part of traditional medicine, to be modernized and to become a kind of science holding the properties being able to do quantitative description and analysis by the way to take in the methods and techniques of quantitative analysis. This is determined by the law, which is true for the developmental process of science from qualitative description to quantitative analysis. Medical science has quickly progressed in this direction. Meanwhile, this progress is also involved in the key step to improve the level of clinical research on TCM at present. To promote this advancing process of quantitative analysis for TCM developing, it is an important step for syndrome differentiation of TCM to be put forward with the abilities to make diagnosis quantitatively. Therefore, this research was aimed at the methodological studies systematically for syndrome differentiation. This is done through inducing in the mathematical model of Bayesian
network system firstly, which was developed recently to analyze complicated events, and on the basis to keep the important property of the concept of holism undisturbed, which is one of the essential principles in the theoretic system of TCM.
This work was composed of following studies.
Firstly, a systematic analysis was made on the methodologies used in the past researches on syndrome differentiation of TCM with quantitative methods. Begun at the modes of thinking for syndrome differentiation, a summary was made to review the application of overall, diverging and imaging thinking modes in the process of syndrome differentiating. Following this, advantages, disadvantages and practical
effectiveness were analyzed for quantitative methods used by various researchers in the study of quantitative syndrome differentiation. As the mathematic basis of quantitative syndrome differentiation, the principles of fuzzy determination, semi-quantitative method, single-factor analysis and multiple-factor analysis have all been applied in this field of study largely. However, no one could be used ideally in the quantitative study of syndrome differentiation, because of their own intrinsic faults. As the developing of principles of artificial intelligence and popularizing of bionic intelligence computer technique, the application of artificial (network) intelligent techniques in the field of syndrome differentiation research should be the new tendency of development.
Secondly, an investigation was carried out on the laws and methods of thinking in syndrome differentiation of TCM. Started at the analysis of correlations among several main methodologies used for syndrome differentiation, syndrome differentiation union system (SDUS) was more carefully discussed based on the principles put forward by Prof. Zhu Wenfong. Here, the key elements of syndrome differentiation were taken as the key link during the thinking process of syndrome differentiation, i.e. lesion location and lesion nature. The effects, position, and its applying in future in the study of quantitative syndrome differentiation were the main subjects absorbing our attention on SDUS. So, the emphasis of this work was put on the further improvement and implementing program for such a SDUS of symptoms-key elements-syndrome, based on calculation of contribution degree for symptoms/signs to the key elements of syndrome differentiation. Particularly, this paper was focused on the combination between quantitative techniques of syndrome differentiation study and SDUS to make it developing further and for it to be used in the research of quantitative syndrome differentiation in TCM.
Thirdly, an investigation was made on the applicability of Bayesian network model used in the quantitative study of TCM syndrome differentiation at the first, based on the analysis as stated above and the review on the newest development in mathematic model of artificial intelligence. The basic consideration for this was on the simulation between Bayesian network model and the th
首先，就中医辨证量化方法学的研究现状进行了系统分析。从辨证思维方法开始，总结了辨证的整体性、发散性、直觉性、形象性思维方法在中医辨证过程中的应用情况，继而分析了迄今为止不同作者在中医辨证量化研究中所用方法的优缺点及其效果。作为辨证量化的数理基础，模糊性判断、半定量方法，多因素分析、人工智能技术均已得到了大量的应用。但是，由于各自的缺陷，无一能理想地运用于中医辨证的量化研究。随着人工智能技术的逐步完善和推广，其在本研究领域的应用已成为新的发展趋势。
其次，对中医辨证思维规律与方法进行了探讨，从分析几种主要辨证方法之间的关系入手，根据导师朱文锋教授所创辨证统一体系理论，确定了辨证思维的关键环节——辨证要素，即病位与病性，讨
论了辨证要素对于建立辨证统一体系的作用及其在辨证定量研究中
的应用前景。因此；在症状与证候名称规范的基础上，探求症状／体
征对辨证要素的贡献度，探讨症状一要素一证侯辨证统一体系的完善
与实施方案，尤其是进一步将人工智能技术与之相结合，以应用于辨
证量化研究。
基于以上分析和人工智能数理模型的新发展，考虑到贝叶斯网
络原理与人脑思维模式，特别是辨证思维过程的良好拟合性，较深入
探讨了该模型应用于中医辨证量化研究的可行性。从网络结构、概率
分析到网络的自学习与经验积累，学习算法及流程图的构建，整个过
程与中医辨证思维与推理具有良好的吻合性。为4b，设计构造了基于
贝叶斯网络的中医量化辨证系统。
为考证该系统的实用效果，本文以一组806例肺系疾病住院病
例资料为样本，分别考察不同样本数贝叶斯网络自学习和预测效果。
结果表明，当样本数达到叩0时，量化辨证的预测就已达到较佳效果，
与人脑辨证思维过程拟合度较好。同时，以此研究样本为基础，分析
了过程中主要相关模块的工作原理与方法。
本研究结果表明，不仅中医辨证量化研究可以收到很好的效果，中医
学体系的量化发展也是可能的。
It is a natural tendency for Traditional Chinese medicine (TCM), as an very important part of traditional medicine, to be modernized and to become a kind of science holding the properties being able to do quantitative description and analysis by the way to take in the methods and techniques of quantitative analysis. This is determined by the law, which is true for the developmental process of science from qualitative description to quantitative analysis. Medical science has quickly progressed in this direction. Meanwhile, this progress is also involved in the key step to improve the level of clinical research on TCM at present. To promote this advancing process of quantitative analysis for TCM developing, it is an important step for syndrome differentiation of TCM to be put forward with the abilities to make diagnosis quantitatively. Therefore, this research was aimed at the methodological studies systematically for syndrome differentiation. This is done through inducing in the mathematical model of Bayesian
network system firstly, which was developed recently to analyze complicated events, and on the basis to keep the important property of the concept of holism undisturbed, which is one of the essential principles in the theoretic system of TCM.
This work was composed of following studies.
Firstly, a systematic analysis was made on the methodologies used in the past researches on syndrome differentiation of TCM with quantitative methods. Begun at the modes of thinking for syndrome differentiation, a summary was made to review the application of overall, diverging and imaging thinking modes in the process of syndrome differentiating. Following this, advantages, disadvantages and practical
effectiveness were analyzed for quantitative methods used by various researchers in the study of quantitative syndrome differentiation. As the mathematic basis of quantitative syndrome differentiation, the principles of fuzzy determination, semi-quantitative method, single-factor analysis and multiple-factor analysis have all been applied in this field of study largely. However, no one could be used ideally in the quantitative study of syndrome differentiation, because of their own intrinsic faults. As the developing of principles of artificial intelligence and popularizing of bionic intelligence computer technique, the application of artificial (network) intelligent techniques in the field of syndrome differentiation research should be the new tendency of development.
Secondly, an investigation was carried out on the laws and methods of thinking in syndrome differentiation of TCM. Started at the analysis of correlations among several main methodologies used for syndrome differentiation, syndrome differentiation union system (SDUS) was more carefully discussed based on the principles put forward by Prof. Zhu Wenfong. Here, the key elements of syndrome differentiation were taken as the key link during the thinking process of syndrome differentiation, i.e. lesion location and lesion nature. The effects, position, and its applying in future in the study of quantitative syndrome differentiation were the main subjects absorbing our attention on SDUS. So, the emphasis of this work was put on the further improvement and implementing program for such a SDUS of symptoms-key elements-syndrome, based on calculation of contribution degree for symptoms/signs to the key elements of syndrome differentiation. Particularly, this paper was focused on the combination between quantitative techniques of syndrome differentiation study and SDUS to make it developing further and for it to be used in the research of quantitative syndrome differentiation in TCM.
Thirdly, an investigation was made on the applicability of Bayesian network model used in the quantitative study of TCM syndrome differentiation at the first, based on the analysis as stated above and the review on the newest development in mathematic model of artificial intelligence. The basic consideration for this was on the simulation between Bayesian network model and the th
引文
第一章
1．梁茂新，等．中医症状量化的方法初探．中国医药学报，1994；9(3)：37
2．樊蔚虹．试论中医证候的分度定量诊断．中医杂志，2000；41(8)：509
3．陆小佐，等．中医计量诊断刍议．中国中医基础医学杂志，1997；3(1)：24
4．贺石林．病证诊断与疗效判断的量化问题．中国医药学报，1991；6(1)：61
5．高汉章，等．中医辨证诊断的数学模型．中国医药学报，1987；2(6)：27
6．侯宗德，等．中西医诊治模式的比较及其互补性．山东中医学院学报，1995；19(3)：158
7．颜文明，等．中医证的研究思路与方法．湖南医药杂志，1982；(4)：56
8．陈家旭．论中医辨证的原则、要求与方法．湖南中医药导报，1995；1(2)：8
9．周慧生．中医模糊诊断方法．中国中医基础医学杂志，1999；5(10)：8
10．王奇，等．中医证候量化的临床流行病学研究初探．广州中医学院学报，1992；9(4)：224
11．梁茂新，等．中医症状量化的方法初探．中国医药学报，1994；9(3)：37
12．赵玉秋，等．流行病学在中医肝证临床辨证标准研究中的应用．中医杂志，1991；(3)：49
13．金益强，等．从肝阳上亢证定量化研究探讨“功能态”．中国人体科学，1992；2(2)：78
14．徐迪华，等．中医量化诊断．南京，江苏科学技术出版社，1997
15．周小青，等．浅析证的等级计量诊断．辽宁中医杂志，1992；(6)：11
16．陈家旭，等．中医计量诊断方法研究进展．中国医药学报，1999；14(6)：63
17．赖世隆，等，中医证候的数理统计基础及血瘀证宏观辨证计量化初探．中国医药学报，1988；3(6)：27
18．潘毅，等．心系疾病心气虚证的心功能指标判别诊断．山东中医杂志，1997；16(4)：151
19．刘士敬，等．中医各系统病证脾气虚证诊断因素的多元逐步回归分析．甘肃中医学院学报，1996；3(1)：9
20．黄益兴，等．头风病证候诊断标准的研究．脑与神经疾病杂志，1997；5(3)：144
21．秦烈，鲍亦万．中医计算机模拟及专家系统概论．北京，人民卫生出版社，1989
22．何振亚．计算智能信息处理．数据采集与处理，1996；11(2)：85
23．杨景灿．中医专家系统的知识表示和推理策略．中国中医药科技，1994；(4)：36
24．朱文锋．建立辨证统一体系之我见．北京中医学院学报，1984；(3)：2
25．杜力晟，赵间忠．中医计算机辅助诊疗系统的临床研究．江苏中医 2000；21(5)：15
26．薛静锋，曹元大．分布式协同中医诊断系统的设计．电脑开发与应用 1999；12(2)：6
27．潘宝宁．对现行中医诊疗系统的探讨．中国中医药信息杂志 1999；6(6)：77
28．马斌荣．《中医专家系统与中医知识库》．北京：北京出版社，1998版
29．王益民，王津生．中医计算机专家系统的建立．天津中医学院学报 1995；14(3)：42
30．杨景灿．中医专家系统的知识表示和推理策略．中国中医药科技 1994；1(4)：36
31．刘自伟．常见内科疾病中医诊疗专家辅助系统的设计及其实现．计算机时代 1994；(4)：1
32．黄卫平．中医专家系统研究概况与展望．医学与哲学 1993；(1)：17
33．秦笃烈．《中医计算机模拟专家系统概论》．北京：人民卫生出版社，1989年版
34．周志坚，毛宗源，邓兆智．神经网络在类风湿性关节炎病情分级中的应用初探．生物医学工程杂志 1999；16(4)：479-482
35．边沁，何裕民，施小成，等．基于MFB-P算法的中医证型的神经网络模型初
探．中国中医基础医学杂志 2001；7(5)：66-69
36．田禾，戴汝为．基于人工神经网强的中医专家系统外壳NNS．计算机学报 2000；(5)：397-400
第二章
37．林求诚．中西医结合对证的研究与展望．中国中医药报 1999；(1)：27
38．门九章，等．中医”证”的研究思路再探讨．中国中医基础医学杂志 1998；4(5)：18
39．陈家旭．中医”证”研究的回顾与展望．北京中医药大学学报 1998；(21)：40
40．任小巧，等．中医证研究方法之我见．中国中医基础医学杂志 1998；4(11)：11
41．马玉宝，等．关于“证”的研究思路及现代表述．中国中医药报 1999；
42．朱文锋，朱咏华．对辨证规律与方法的研究．湖南中医学院学报 2002；22(2)：1
43．梁伟雄．中医辨证定量化的思路．广州中医药大学学报 1997；14(1)：5-7
44．朱文锋．辨证统一体系的创立．中国中医基础医学杂志 2001；7(4)：4
45．方药中．辨证论治七讲[M]．北京：人民卫生出版社，1978年月12月一版
46．朱文锋．中医病证规范化之研究．中国医药学报 1996；11(5)：4-6．
47．国家技术监督管理局．中医临床诊疗术语-证候部分，北京：学苑出版社，1999；
48．黎敬波．建立临床证候诊断标准的基本原则和方法．中国中医药报 2000；(5)3
49．王建华 脾气虚证本质研究的途径及其方法．中医杂志 1998；39(1)：50
50．尹必武．证候临床诊断标准规范刍议．中国医药学报 2000；15(3)：6-9
51．吴圣贤，等．中医计量诊断研究概况．北京中医药大学学报 1998；21(3)：35
52．陈辉，等．中医气虚证量化指标初探．中国中西医结合杂志 1998；18(7)：391
53．陆小佐，等．中医计量诊断刍议．中国中医基础医学杂志 1997；3(1)：24
54．潘毅，等．心系疾病常见证型的计量诊断．中国中医药学会建会20周年学术
年会专辑，415(3)：15
55．朱咏华，朱文锋，等．常见症状的计量辨证．辽宁中医杂志 2000；27(6)-(12)：243
56．张震．证候探微．北京中医学院学报 1984；7(5)：2-7．
57．黄星垣．中医病证规范的层次和框架．中国医药学报 1990；5(4)：3-5
58．欧阳琦．中医病证症三联诊疗．北京：人民卫生出版社，1998年1月一版
59．宋镇星．中医证本质的研究方法与思路．中国中医药学报 2002；17(3)：179-182
第三章
60．林士敏，田凤占，陆玉昌．贝叶斯网络的建造及其在数据采掘中的应用．清华大学学报(自然科学版)，2001；41(1)：49-50
61．慕春棣，戴剑彬 叶俊，用于数据挖掘的贝叶斯网络，软件学报 2000；11(5)：660～666
62. Spirtes, P., Glymour, C. and Scheines, R. 1993. Causation, Prediction and Search. Springer-Verlag.
63. Cheng, J. and Greiner, R. 1999. comparing Bayesian network classifiers,UAI99 pp 101-108.
64. Cheng, J. and Greiner, R. 2001. Learning Bayesian Belief Network Classifiers: Algorithms and System. Canadian Conference on AI 2001:141-151.
65. Domingos, P., and Pazzani, M. 1997. Beyond independence: Conditions for the optimality of the simple Bayesian classifier. Machine Learning 29:103—130.
66. Duda, R. and Hart, P. 1973. Pattern classification and scene analysis.
John Wiley & Sons.
67． Friedman, N., Geiger, D. and Goldszmidt, M. 1997． Bayesian Network Classifiers. Machine Learning, 29, pp. 131-161．
68． Greiner, R. Grove, A. and Schuurmans, D. 1997． Learning Bayesian nets that perform well. In Proceedings of UAI-97．
69． Kohavi, R. and John, G. 1997． Wrappers for feature subset selection. In Artificial Intelligence journal, special issue on relevance, vol.97, No. 1-2, pp273-324．
70． Kononenko, I. 1991． Semi-naive Bayesian classifier. In Y. Kodratoff (ed.), Proceedings of sixth European working session on learning. Springer-Verlag. pp206-219．
71． Langley, P., Iba, W. and Thompson, K. 1992． An analysis of Bayesian classifiers. In Proceedings of 风热犯肺 AI-92． pp223-228．
72． Langey, P. and Sage, S. 1994． Induction of selective Bayesian classifiers. In Proceedings of UAI-94．
73． Murphy, P.M. and Aha, D. W. 1995． UCI repository of machine learning databases, http://www.ics.uci.edu/ mlearn/MLRepository. html.
74． Pazzani, M.J. 1995． Searching for dependencies in Bayesian classifiers. In Proceedings of AI & STAT'95．
75． 王军，周伟达，贝叶斯网络的研究与进展，电子科技 1999; (8) : 22
76． Singh, M. and Provan, G. M. 1996． Efficient learning of selective Bayesian network classifiers. In Proceedings of the ICML-96．
77． 董雁适，程翼宇，潘云鹤，基于因果关系发现的关键化学组分辨识方法，等发表.
1．梁茂新，等．中医症状量化的方法初探．中国医药学报，1994；9(3)：37
2．樊蔚虹．试论中医证候的分度定量诊断．中医杂志，2000；41(8)：509
3．陆小佐，等．中医计量诊断刍议．中国中医基础医学杂志，1997；3(1)：24
4．贺石林．病证诊断与疗效判断的量化问题．中国医药学报，1991；6(1)：61
5．高汉章，等．中医辨证诊断的数学模型．中国医药学报，1987；2(6)：27
6．侯宗德，等．中西医诊治模式的比较及其互补性．山东中医学院学报，1995；19(3)：158
7．颜文明，等．中医证的研究思路与方法．湖南医药杂志，1982；(4)：56
8．陈家旭．论中医辨证的原则、要求与方法．湖南中医药导报，1995；1(2)：8
9．周慧生．中医模糊诊断方法．中国中医基础医学杂志，1999；5(10)：8
10．王奇，等．中医证候量化的临床流行病学研究初探．广州中医学院学报，1992；9(4)：224
11．梁茂新，等．中医症状量化的方法初探．中国医药学报，1994；9(3)：37
12．赵玉秋，等．流行病学在中医肝证临床辨证标准研究中的应用．中医杂志，1991；(3)：49
13．金益强，等．从肝阳上亢证定量化研究探讨“功能态”．中国人体科学，1992；2(2)：78
14．徐迪华，等．中医量化诊断．南京，江苏科学技术出版社，1997
15．周小青，等．浅析证的等级计量诊断．辽宁中医杂志，1992；(6)：11
16．陈家旭，等．中医计量诊断方法研究进展．中国医药学报，1999；14(6)：63
17．赖世隆，等，中医证候的数理统计基础及血瘀证宏观辨证计量化初探．中国医药学报，1988；3(6)：27
18．潘毅，等．心系疾病心气虚证的心功能指标判别诊断．山东中医杂志，1997；16(4)：151
19．刘士敬，等．中医各系统病证脾气虚证诊断因素的多元逐步回归分析．甘肃中医学院学报，1996；3(1)：9
20．黄益兴，等．头风病证候诊断标准的研究．脑与神经疾病杂志，1997；5(3)：144
21．秦烈，鲍亦万．中医计算机模拟及专家系统概论．北京，人民卫生出版社，1989
22．何振亚．计算智能信息处理．数据采集与处理，1996；11(2)：85
23．杨景灿．中医专家系统的知识表示和推理策略．中国中医药科技，1994；(4)：36
24．朱文锋．建立辨证统一体系之我见．北京中医学院学报，1984；(3)：2
25．杜力晟，赵间忠．中医计算机辅助诊疗系统的临床研究．江苏中医 2000；21(5)：15
26．薛静锋，曹元大．分布式协同中医诊断系统的设计．电脑开发与应用 1999；12(2)：6
27．潘宝宁．对现行中医诊疗系统的探讨．中国中医药信息杂志 1999；6(6)：77
28．马斌荣．《中医专家系统与中医知识库》．北京：北京出版社，1998版
29．王益民，王津生．中医计算机专家系统的建立．天津中医学院学报 1995；14(3)：42
30．杨景灿．中医专家系统的知识表示和推理策略．中国中医药科技 1994；1(4)：36
31．刘自伟．常见内科疾病中医诊疗专家辅助系统的设计及其实现．计算机时代 1994；(4)：1
32．黄卫平．中医专家系统研究概况与展望．医学与哲学 1993；(1)：17
33．秦笃烈．《中医计算机模拟专家系统概论》．北京：人民卫生出版社，1989年版
34．周志坚，毛宗源，邓兆智．神经网络在类风湿性关节炎病情分级中的应用初探．生物医学工程杂志 1999；16(4)：479-482
35．边沁，何裕民，施小成，等．基于MFB-P算法的中医证型的神经网络模型初
探．中国中医基础医学杂志 2001；7(5)：66-69
36．田禾，戴汝为．基于人工神经网强的中医专家系统外壳NNS．计算机学报 2000；(5)：397-400
第二章
37．林求诚．中西医结合对证的研究与展望．中国中医药报 1999；(1)：27
38．门九章，等．中医”证”的研究思路再探讨．中国中医基础医学杂志 1998；4(5)：18
39．陈家旭．中医”证”研究的回顾与展望．北京中医药大学学报 1998；(21)：40
40．任小巧，等．中医证研究方法之我见．中国中医基础医学杂志 1998；4(11)：11
41．马玉宝，等．关于“证”的研究思路及现代表述．中国中医药报 1999；
42．朱文锋，朱咏华．对辨证规律与方法的研究．湖南中医学院学报 2002；22(2)：1
43．梁伟雄．中医辨证定量化的思路．广州中医药大学学报 1997；14(1)：5-7
44．朱文锋．辨证统一体系的创立．中国中医基础医学杂志 2001；7(4)：4
45．方药中．辨证论治七讲[M]．北京：人民卫生出版社，1978年月12月一版
46．朱文锋．中医病证规范化之研究．中国医药学报 1996；11(5)：4-6．
47．国家技术监督管理局．中医临床诊疗术语-证候部分，北京：学苑出版社，1999；
48．黎敬波．建立临床证候诊断标准的基本原则和方法．中国中医药报 2000；(5)3
49．王建华 脾气虚证本质研究的途径及其方法．中医杂志 1998；39(1)：50
50．尹必武．证候临床诊断标准规范刍议．中国医药学报 2000；15(3)：6-9
51．吴圣贤，等．中医计量诊断研究概况．北京中医药大学学报 1998；21(3)：35
52．陈辉，等．中医气虚证量化指标初探．中国中西医结合杂志 1998；18(7)：391
53．陆小佐，等．中医计量诊断刍议．中国中医基础医学杂志 1997；3(1)：24
54．潘毅，等．心系疾病常见证型的计量诊断．中国中医药学会建会20周年学术
年会专辑，415(3)：15
55．朱咏华，朱文锋，等．常见症状的计量辨证．辽宁中医杂志 2000；27(6)-(12)：243
56．张震．证候探微．北京中医学院学报 1984；7(5)：2-7．
57．黄星垣．中医病证规范的层次和框架．中国医药学报 1990；5(4)：3-5
58．欧阳琦．中医病证症三联诊疗．北京：人民卫生出版社，1998年1月一版
59．宋镇星．中医证本质的研究方法与思路．中国中医药学报 2002；17(3)：179-182
第三章
60．林士敏，田凤占，陆玉昌．贝叶斯网络的建造及其在数据采掘中的应用．清华大学学报(自然科学版)，2001；41(1)：49-50
61．慕春棣，戴剑彬 叶俊，用于数据挖掘的贝叶斯网络，软件学报 2000；11(5)：660～666
62. Spirtes, P., Glymour, C. and Scheines, R. 1993. Causation, Prediction and Search. Springer-Verlag.
63. Cheng, J. and Greiner, R. 1999. comparing Bayesian network classifiers,UAI99 pp 101-108.
64. Cheng, J. and Greiner, R. 2001. Learning Bayesian Belief Network Classifiers: Algorithms and System. Canadian Conference on AI 2001:141-151.
65. Domingos, P., and Pazzani, M. 1997. Beyond independence: Conditions for the optimality of the simple Bayesian classifier. Machine Learning 29:103—130.
66. Duda, R. and Hart, P. 1973. Pattern classification and scene analysis.
John Wiley & Sons.
67． Friedman, N., Geiger, D. and Goldszmidt, M. 1997． Bayesian Network Classifiers. Machine Learning, 29, pp. 131-161．
68． Greiner, R. Grove, A. and Schuurmans, D. 1997． Learning Bayesian nets that perform well. In Proceedings of UAI-97．
69． Kohavi, R. and John, G. 1997． Wrappers for feature subset selection. In Artificial Intelligence journal, special issue on relevance, vol.97, No. 1-2, pp273-324．
70． Kononenko, I. 1991． Semi-naive Bayesian classifier. In Y. Kodratoff (ed.), Proceedings of sixth European working session on learning. Springer-Verlag. pp206-219．
71． Langley, P., Iba, W. and Thompson, K. 1992． An analysis of Bayesian classifiers. In Proceedings of 风热犯肺 AI-92． pp223-228．
72． Langey, P. and Sage, S. 1994． Induction of selective Bayesian classifiers. In Proceedings of UAI-94．
73． Murphy, P.M. and Aha, D. W. 1995． UCI repository of machine learning databases, http://www.ics.uci.edu/ mlearn/MLRepository. html.
74． Pazzani, M.J. 1995． Searching for dependencies in Bayesian classifiers. In Proceedings of AI & STAT'95．
75． 王军，周伟达，贝叶斯网络的研究与进展，电子科技 1999; (8) : 22
76． Singh, M. and Provan, G. M. 1996． Efficient learning of selective Bayesian network classifiers. In Proceedings of the ICML-96．
77． 董雁适，程翼宇，潘云鹤，基于因果关系发现的关键化学组分辨识方法，等发表.