A Formal Model for Experience–Aided Diagnosis

This paper presents a novel approach to model–based diagnosis. The approach addresses the two main problems that have prevented model–based diagnostic techniques from being widely used: computational complexity of abduction and inadequacies of device models. A model for automated diagnosis is defined that combines (1) deduction to rule out hypotheses, (2) abduction to generate hypotheses, and (3) induction to recall past experiences and account for potential errors in the device models. A review of the three forms of inference is provided, as well as a detailed analysis of the relationship between case–based reasoning and induction. The proposed model for diagnosis is used to characterize diagnostic errors and relate them to different types of errors in the device models. Experimental results are then described and used to assert the practicality and the usefulness of the approach. The model presented in this paper yields a practical method for solving hard diagnostic problems at a reasonable computational cost and provides a theoretical basis for overcoming the problem of partially incorrect device models.     

Integrating integer programming and probabilistic deduction graphs for probabilistic reasoning

Optimal solutions of several variants of the probabilistic reasoning problem were found by a new technique that integrates integer programming and probabilistic deduction graphs (PDG). PDGs are extended from deduction graphs of the and-type via normal deduction graphs. The foregoing variants to be solved can involve multiple hypotheses and multiple evidences where the former is given and the latter is unknown and being found or vice versa. The relationship among these hypotheses and evidences with possible intermediaries is represented by a causal graph. The proposed method can handle a large causal graph of any type and find an optimal solution by invoking a linear integer programming package. In addition, formulating the reasoning problem to fit integer programming takes a polynomial time. H.-L. Li was visiting the Department of Computer Sciences, University of North Texas in 1988–1989. He is with the Institute of Information Management, National Chiao Tung University, Hsinchu, Taiwan, R.O.C.     

农用柴油机故障诊断系统的研究与开发

根据小型拖拉机的常见故障,利用Microsoft Visual FoxPro编程工具和人工智能专家系统原理,建立了知识库和相应知识表达推理机制,设计并组建了农用柴油发动机故障诊断专家系统,缓解了故障诊断专家供不应求的矛盾,提高了农用柴油机的常见故障诊断的效率和准确率。该系统主要实现的功能包括:发动机故障诊断;故障模糊查询;用户诊断数据库浏览、打印;知识库维护修改和扩充。     

Acquiring and representing strategic knowledge in the diagnosis domain

Abstract: Techniques for acquiring and representing strategic knowledge for guiding diagnostic processes are presented. In a diagnostic expert system, strategic knowledge can be represented either by a specific knowledge base or it can be 'embedded' into the inference engine. We decided for the former; so that knowledge can be acquired or modified without affecting the problem solving paradigm. Strategic knowledge is acquired by expert interview in a straightforward way: on the basis of simple information provided by the expert, an internal sophisticated representation is automatically generated. The techniques are not restricted to a particular problem-solving paradigm or application. However, in order to prove the effectiveness of our approach, a problem solving paradigm is also presented. The paradigms adopted in diagnosis must face two problems: the selection of the 'right' hypothesis (fault) to pursue and the selection of the 'right' observation (measurement) to be executed. We present some criteria for selecting hypotheses and observations. Our proposal is suitable for domains where the measurements to localise the fault do not always provide certainty but only a 'degree of belief' about the presence of the fault. As a consequence, the problem of selecting the right measurement is solved by appropriate criteria and heuristic reasoning. Moreover, we do not consider 'right' as a predefined concept: actually, it is based on the information provided by the expert. So he can define this concept on the basis of his own judgment.     

Non-monotonic characterization of induction and its application to inductive learning

In this article a new approach to the formalization of inductive inference in terms of non-monotonic inference is proposed. Induction is characterized as closed-world reasoning from the available data, followed by an inductive jump, which consists in assuming that valid conclusions in the database (assuming closed-world) hold also in the rest of the world. This conception of induction results is adequate to characterize those inference processes that could be formalized, that is, those based in analytical procedures of pattern-matching or regularity detection in the available data. the proposed characterization formally describes the implicit deductive processes of induction and its non-monotonic nature, and could be used as an abstract model of the mental process that leads to obtaining inductive hypotheses. This proposal reduces the problem of induction automatization to that of deduction automatization. Also, it constitutes a formal framework that covers several inductive inference methods used in machine learning. Besides it formalizes inductive definitions, which are very common in science and computer science. © 1995 John Wiley & Sons, Inc.     

溯因推理问题的碰集求解方法

溯因推理为归纳与演绎推理之外的另一种重要的推理形式,在人工智能等领域有着广泛的应用.通俗地讲,溯因推理是从观察(结果)去推断原因的推理过程.不同于以往的研究思路,通过使用本原蕴含式和素蕴含,证明了可以把命题逻辑和命题模态逻辑系统S5中求溯因问题的极小解释转化为求对应集合的极小碰集问题.给出了求解溯因问题的一种新方法.     

A fuzzy expert database system

Fuzzy concepts always exist in much of human reasoning as well as decision making. This paper presents a fuzzy expert database system which is an integration of a fuzzy expert system building tool called SYSTEM Z-II and a database management system called Rdb/VMS. This system is able to extract fuzzy data and terms stored in a database and used in the fuzzy reasoning in an expert system. It can also retrieve information by fuzzy database-queries which are generated by the expert system automatically. Many expert systems in different domain areas such as decision making can be constructed. Sample applications are described to demonstrate the flexibility and power of this system. The fuzzy query language defined and used in the system can also be used independently as a fuzzy enquiry tool in database applications.     

针对教学场景的ZFC集合论Coq形式化

离散数学是计算机类专业的基础课程之一,命题逻辑、一阶逻辑与公理集合论是其重要组成部分.教学实践表明,初学者准确理解语法、语义、推理系统等抽象概念是有一定难度的.近年来,已有一些学者开始在教学中引入交互式定理证明工具,以帮助学生构造形式化证明,更透彻地理解逻辑系统.然而,现有的定理证明器有较高上手门槛,直接使用会增加学生的学习负担.鉴于此,在Coq中开发了针对教学场景的ZFC公理集合论证明器.首先,形式化了一阶逻辑推理系统和ZFC公理集合论;之后,开发了数条自动化推理规则证明策略.学生可以在与教科书风格相同的简洁证明环境中使用自动化证明策略完成定理的形式化证明.该工具被用在了大一新生离散数学课程的教学中,没有定理证明经验的学生使用该工具可以快速完成数学归纳法和皮亚诺算术系统等定理的形式化证明,验证了该工具的实际效果.     

Developing an expert law tutor

This paper outlines the development considerations which led to the construction of a prototype “expert law tutor”. This is a system which models the competencies of the trained law teacher and is intended primarily for use as an interactive student‐directed multidimensional learning tool. The aim is to computerize those activities which might be encountered in a real‐life tutorial exercise—research, reasoning, problem solving and advising, testing, diagnosis and feedback. The paper takes a liberal view of what constitutes an expert system. The terminology expert system is retained but is defined broadly, perhaps contrary to convention, to encompass not only an advisory component or a diagnostic tool but other tutorial functions. To highlight this difference reference is made to an “expert law tutor”. It is the “expertise” of the human law tutor which is called upon. For this reason the system includes a hypertext library or information component and an multiple choice assessment component as well as a legal adviser. Providing “expert” tutor feedback is a key element in the dialogue between system and student. The prototype system was developed using (inter alia) a rule‐based expert system shell and an object‐oriented hypertext tool, and the paper comments on the features of these development tools.     

Machine Learning and the foundations of inductive inference

The problem of valid induction could be stated as follows: are we justified in accepting a given hypothesis on the basis of observations that frequently confirm it? The present paper argues that this question is relevant for the understanding of Machine Learning, but insufficient. Recent research in inductive reasoning has prompted another, more fundamental question: there is not just one given rule to be tested, there are a large number of possible rules, and many of these are somehow confirmed by the data — how are we to restrict the space of inductive hypotheses and choose effectively some rules that will probably perform well on future examples? We analyze if and how this problem is approached in standard accounts of induction and show the difficulties that are present. Finally, we suggest that the explanation-based learning approach and related methods of knowledge intensive induction could be, if not a solution, at least a tool for solving some of these problems.     

NFI: a neuro-fuzzy inference method for transductive reasoning

This paper introduces a novel neural fuzzy inference method-NFI for transductive reasoning systems. NFI develops further some ideas from DENFIS-dynamic neuro-fuzzy inference systems for both online and offline time series prediction tasks. While inductive reasoning is concerned with the development of a model (a function) to approximate data in the whole problem space (induction), and consecutively-using this model to predict output values for a new input vector (deduction), in transductive reasoning systems a local model is developed for every new input vector, based on some closest to this vector data from an existing database (also generated from an existing model). NFI is compared with both inductive connectionist systems (e.g., MLP, DENFIS) and transductive reasoning systems (e.g., K-NN) on three case study prediction/identification problems. The first one is a prediction task on Mackey Glass time series; the second one is a classification on Iris data; and the last one is a real medical decision support problem of estimating the level of renal function of a patient, based on measured clinical parameters for the purpose of their personalised treatment. The case studies have demonstrated better accuracy obtained with the use of the NFI transductive reasoning in comparison with the inductive reasoning systems.     

A knowledge-based control architecture with interactive reasoningfunctions

A knowledge-based system architecture called IPEX is presented that uses a time-distributed, interactive reasoning paradigm for process control applications. Structural features of the system are presented, and it is shown that temporal considerations are included in each of the system's data structures either explicitly or implicitly. Diagnostic planning is discussed, and explanations of the algorithms that formulate and maintain diagnostic/control plans are given. In particular, it is shown that the IPEX system can manage concurrently executing interactive diagnostic plans for multiple problem hypotheses     

基于多级推理的玉米病虫害专家系统

详细介绍了玉米病虫害专家系统的设计与开发,包括数据库,知识库以及推理机的设计,针对玉米病虫害多样性增加导致传统的专家系统诊断时出现的同一症状对应多个诊断结果的粗糙诊断问题,提出基于多级推理的玉米病虫害专家系统,根据玉米的发病症状,在病虫害的可能性中进行综合判断,以发病时期,发病条件,发病部位为依据采用多级推理机制,以降低推理结果集元素的数量,同时可视知识参与推理,为推理提供感观信息,提高推理的准确性。     

基于ESTA构建逆向推理专家系统

介绍专家系统工具ESTA和逆向推理机制的基本概念,阐述使用ESTA构建逆向推理专家系统的基本方法。     

INTEGRATING CLASSIFICATION-BASED COMPILED LEVEL REASONING WITH FUNCTION-BASED DEEP LEVEL REASONING

Problem solving based on compiled associations between elements of the decision space and data is an efficient mode of reasoning for a large percentage of situations faced by an expert. But in some (usually small) percentage of cases, compiled associations are not enough by themselves to lead to correct results. Reasoning from “deeper” levels of understanding offers the advantage of producing correct results even in atypical cases, but at the cost of expanding more computational resources. Thus the trade-off between compiled level systems and deep level systems is between computational efficiency (at the compiled level) and problem-solving generality (at the deep level). We describe a hybrid system containing elements of both deep level reasoning and compiled level reasoning. More particularly, we propose a problem-solving architecture for category-based diagnostic problem solving which at the compiled level centers on classification problem solving and at the deep level uses a type of function-based reasoning. We concentrate in this report on the interaction between the compiled and deep level units and on the mechanisms of function-based reasoning that we employ. We show how our function-based consequence-finding problem solver can be focused by problem solving at the compiled level and how, through such interaction, we obtain the computational efficiency characteristic of compiled level problem solving while retaining the robustness characteristic of deep level problem solving.     

推理技术在决策支持系统中的应用

文中首先阐述了专家系统和不精确推理的基本理论，随后讨论了不精确推理技术在Fuzzy CLIPS中的实现，最后用实例说明如何在决策支持系统中应用专家系统的推理能力。文中提出了一套将专家系统推理机、知识库与数据库应用程序结合起来以解决复杂推理问题的方法。这种方法具有较强的通用性，稍作修改即可用于在不同平台上解决不同领域的问题。     

ENHANCEMENT OF THE CAPABILITIES OF AN NMR IMAGING MACHINE USING EXPERT SYSTEMS

ABSTRACT

Practical effectiveness of NMR imaging in diagnostic medicine can be considerably upgraded by incorporating into the machine high-level intelligent software support. Partly because NMR imaging is a relatively new technology, knowledge acquisition is essentially related to incoming new experience. Therefore an expert system approach to NMR medical applications should rely on rule induction techniques based on a series of example expert decisions. The complete project consists of three main components: (1) a protocol expert system, (2) a diagnosis expert system, and (3) a vision system. Expert system prototypes regarding part 1 and 2 of this study were built indicating preliminary interesting results. These results justify our attempts aimed at the enhancement of NMR capabilities as a diagnostic tool and consequent commercial benefits.     

An integrated case-based reasoning approach for intelligent help desk fault management

As modern business functions become more complex and knowledge-intensive, with increasing demands for quality services, there is an emerging trend for organisations to develop and deploy intelligent knowledge-based systems for mission-critical operations. Some of the challenges in successfully implementing this breed of systems depend on how well the intelligent system is integrated with conventional existing information systems and workflow, and the quality of the intelligent system itself. Developing quality expert systems lies in the effective modelling of cognitive processes of human experts and representation of various forms of related knowledge in a domain. An integrated intelligent system called the Intelligent Help Desk Facilitator (IHDF), has been developed for computer and network fault management. The system, which comprises various modules including an expert system, is successfully deployed in a problem response help desk environment of a local bank. This paper describes a cognitive-driven approach to the development of the expert system based on a hybrid knowledge representation and reasoning strategy. The approach incorporates a hybrid case-based reasoning (CBR) framework of techniques which include case memory organisation structures (discrimination networks and shared-featured networks), case indexing and retrieval schemes (fuzzy character-matching, nearest-neighbour similarity matching and knowledge-guided indexing); and an interactive and incremental style of reasoning. The paper discusses the design and implementation of the expert system component of IHDF and illustrates the appropriateness of the hybrid architecture for problem resolution and diagnostic types of applications.     

航电设备故障诊断专家系统

故障诊断专家系统在自动测试与诊断领域有着广泛的应用,它是自动测试系统的核心技术之一．介绍了某型机航电设备故障诊断专家系统,运用规则推理与Hash算法相结合的综合推理方式进行故障快速定位及故障预测,提出视情维修建议,提高了航电系统的故障检测率及故障诊断效率．     

基于FPN—RSVM的电梯故障诊断方法

模糊Petri网(FPN)作为产生式规则系统的建模、知识表示和诊断推理的工具,在故障诊断领域得到了广泛的应用。但调整学习机制的缺乏使其不能处理专家系统的变化。使用回归型支持向量机(RSVM)对电梯在不同运行状态下的专家诊断数据进行训练,可以得到与电梯状态相对应的FPN诊断网络权值。故障发生时,根据电梯运行状态选择相应的权值,可以使得诊断结果更加准确,更加符合实际情况。