No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
Abstract
A new approach for uncertain causality representation and probabilistic reasoning named as dynamic uncertain causality graph (DUCG) was presented previously, in which only the discrete variables and certain evidence were addressed. In this paper, the free mixtures of discrete and continuous variables as well as uncertain evidence are addressed. The general idea to deal with continuous variables is to transform them into fuzzy discrete variables along with their corresponding uncertain causalities, and then treat them as ordinary discrete variables, which involves how to deal with fuzzy evidence. It is pointed out that uncertain evidence is either: 1) fuzzy evidence that is an observed certain value of a continuous variable falling into a fuzzy area across two or more fuzzy discrete states of a variable or 2) soft evidence that can only be understood as a probability distribution over the states of a variable. The algorithm for utilizing uncertain evidence in inference is presented, in which uncertain evidence is treated as a virtual child variable of the observed variable without changing the knowledge and inference algorithm encoded in DUCG. It is proved that the two types of uncertain evidence are the same in nature and can be treated indiscriminatingly. Moreover, this method dealing with fuzzy evidence in DUCG can be used for failure forecasting of systems. Examples are provided to illustrate the methodology.
... As a newly graphical model, dynamic uncertain causality graph model(DUCG) can address the drawback of requirement of complete information. DUCG model is proposed by Li Li lily122008@163.com 1 Central South University, Changsha 410083, China Zhang [8,9] for knowledge representation and reasoning. DUCG links events with causality chain by graphic symbols [10] definitely and easily. ...
... DUCG provides an effective way of knowledge representation and inference to the causality relation among observable variables. As shown in Fig. 1, a typical DUCG with directed cyclic cases uses different type of event to express the variables and the uncertain causality links between them [8,9]. The types are defined as F-, B-, X-in Fig. 1. ...
... It is the same with the states of B 12 ,· · · , B 1k and other variables. Definition 1 [8,9] Suppose variable V is either Xor Btype variable, in the case of matrices, the basic model of DUCG is defined as ...
Dynamic uncertain causality graph (DUCG), which is based on probability theory, is used for uncertain knowledge representation and reasoning. However, the traditional DUCG has difficulty expressing the causality of the events with crisp numbers. Therefore, an intuitionistic fuzzy set based dynamic uncertain causality graph (IFDUCG) model is proposed in this paper. The model focuses on describing the uncertain event in the form of intuitionistic fuzzy sets, which can handle with the problem of describing vagueness and uncertainty of an event in the traditional model. Then the technique for order preference by similarity to an ideal solution (TOPSIS) method is combined with IFDUCG for knowledge representation and reasoning so as to integrate more abundant experienced knowledge into the model to make the model more reliable. Then some examples are used to validate the proposed method. The experimental results prove that the proposed method is effective and flexible in dealing with the difficulty of the fuzzy event of knowledge representation and reasoning. Furthermore, we make a practical application to root cause analysis of aluminum electrolysis and the results show that the proposed method is available for workers to make decisions.
... It is of solid theoretical foundation, with directed acyclic graph to express causal dependency and conditional probability table to quantify the uncertainty of causality. To improve the efficiency of FD, a graphical inference methodology named Dynamic Uncertain Causality Graph (DUCG) has been developed and successfully tested in many practical cases [44][45][46][47][48][49][50][51]. ...
... and variable X n,k in state k, where i = n and j = k acyclic graph or a directed cyclic graph, with a probabilistic-based or evidence-based reasoning [44,45]. Nevertheless, the causal graphs construction of DUCG is complicate in some cases, turning out redundant on the symptoms needed by DUCG to correctly classify the occurring faults [46][47][48][49][50][51]. In this paper, we propose a method to simplify the inference rules of DUCG based on a genetic algorithm (GA) that uses an FDT as surrogate model of the DUCG. ...
... where A n,k;i,j accounts for the mechanism that V i,j induces on X n,k , without considering any other natural interaction with other mutual variables B i,j , j 0 = j; and (r n;i /r n ) is a weighting factor of A n,k;i,j , with r n P i r n;i and r n;i being the causal relationship intensity between each V i and X n [45][46][47][48]. ...
Fault diagnostics is important for safe operation of nuclear power plants (NPPs). In recent years, data-driven approaches have been proposed and implemented to tackle the problem, e.g., neural networks, fuzzy and neuro-fuzzy approaches, support vector machine, K-nearest neighbor classifiers and inference methodologies. Among these methods, dynamic uncertain causality graph (DUCG) has been proved effective in many practical cases. However, the causal graph construction behind the DUCG is complicate and, in many cases, results redundant on the symptoms needed to correctly classify the fault. In this paper, we propose a method to simplify causal graph construction in an automatic way. The method consists in transforming the expert knowledge-based DCUG into a fuzzy decision tree (FDT) by extracting from the DUCG a fuzzy rule base that resumes the used symptoms at the basis of the FDT. Genetic algorithm (GA) is, then, used for the optimization of the FDT, by performing a wrapper search around the FDT: the set of symptoms selected during the iterative search are taken as the best set of symptoms for the diagnosis of the faults that can occur in the system. The effectiveness of the approach is shown with respect to a DUCG model initially built to diagnose 23 faults originally using 262 symptoms of Unit-1 in the Ningde NPP of the China Guangdong Nuclear Power Corporation. The results show that the FDT, with GA-optimized symptoms and diagnosis strategy, can drive the construction of DUCG and lower the computational burden without loss of accuracy in diagnosis.
... Dynamic uncertain causality graph (DUCG) is newly presented in [21]- [27] to deal with large and complex systems with dynamics and uncertainties. The fundamental researches of DUCG have been done: basic methodology in [22] and [23], statistic base and matrix model in [24], methodology dealing with directed cyclic graphs in [25], methodology for dynamic diagnosis in [26], and methodology dealing with continuous variables, uncertain evidence, and failure forecast in [27]. ...
... Dynamic uncertain causality graph (DUCG) is newly presented in [21]- [27] to deal with large and complex systems with dynamics and uncertainties. The fundamental researches of DUCG have been done: basic methodology in [22] and [23], statistic base and matrix model in [24], methodology dealing with directed cyclic graphs in [25], methodology for dynamic diagnosis in [26], and methodology dealing with continuous variables, uncertain evidence, and failure forecast in [27]. DUCG has been applied in large and complex industrial systems [28]- [30], but DUCG has not been applied to reliability. ...
... DUCG proposed by Zhang et al. [22]- [27] is a graphical model of intelligent systems dealing with uncertain causal knowledge representation and probabilistic reasoning for many intelligent issues such as fault diagnoses of large and complex [26], DUCG consists of a set of variables or events that can be classified as B-type, X-type, D-type, and G-type nodes, and directed arcs containing P-type or F-type events. ...
Probabilistic safety assessment (PSA) has been widely applied to large complex industrial systems like nuclear power plants, chemical plants, etc. Event trees (ET) and fault trees (FT) are the major tools, but dependences and logic cycles may exist among and within them, and are not well addressed, leading to even optimistic estimates. Repeated representations and calculations exist. Causalities are assumed deterministic, while sometimes they are uncertain. This paper applies dynamic uncertain causality graph (DUCG) in PSA to overcome these problems. DUCG is a newly presented approach for uncertain causality representation and probabilistic reasoning, and has been successfully applied to online fault diagnoses of large complex industrial systems. This paper suggests to model all ETs and FTs of a target system as a single DUCG allowing uncertain causalities and avoiding repeated representations, and calculate the probabilities/frequencies of the undesired events by using the DUCG algorithm. In the calculation, the problems of dependencies and circular loops are solved. The suggested DUCG representation mode and calculation algorithm are presented and illustrated with examples. The results reveal the effectiveness and feasibility of this methodology.
... Yu et al. [17] proposed the networked process monitoring framework based on DBN for fault detection and root cause diagnosis. Zhang [18] designed a dynamic uncertain causality graph based on DBN for knowledge representation and probabilistic reasoning to implement interpretable causal reasoning. Furthermore, dividing complex industrial processes into blocks is a widely adopted technique for integrating domain knowledge in root cause analysis. ...
With the development of intelligent manufacturing and the increasing complexity of industrial production, root cause diagnosis has gradually become an important research direction in the field of industrial fault diagnosis. However, existing research methods struggle to effectively combine domain knowledge and industrial data, failing to provide accurate, online, and reliable root cause diagnosis results for industrial processes. To address these issues, a novel fault root cause diagnosis framework based on knowledge graph and industrial data, called Root-KGD, is proposed. Root-KGD uses the knowledge graph to represent domain knowledge and employs data-driven modeling to extract fault features from industrial data. It then combines the knowledge graph and data features to perform knowledge graph reasoning for root cause identification. The performance of the proposed method is validated using two industrial process cases, Tennessee Eastman Process (TEP) and Multiphase Flow Facility (MFF). Compared to existing methods, Root-KGD not only gives more accurate root cause variable diagnosis results but also provides interpretable fault-related information by locating faults to corresponding physical entities in knowledge graph (such as devices and streams). In addition, combined with its lightweight nature, Root-KGD is more effective in online industrial applications.
... When we need to modify the DUCG model, we only need to modify the local knowledge in the corresponding sub-DUCGs, to achieve the purpose of modifying the whole DUCG model. Some other features of DUCG include: (1) DUCG can deal with loops, so the DUCG model supports the expression of causal loops [26]; (2) DUCG can deal with discrete, continuous, and fuzzy evidence, which increases the robustness of the model [25,27]; (3) the causal reasoning of DUCG depends much on the structure of the model and has low requirements for the precision of model parameters; (4) DUCG can realize the concise expression of knowledge and allow the incomplete expression of knowledge. In DUCG, the causal mechanism between a child variable and its parent variables is shown in Figure 2. The child event Xnk may be caused by one or more parent events. ...
The causes of sore throat are complex. It can be caused by diseases of the pharynx, adjacent organs of the pharynx, or even systemic diseases. Therefore, a lack of medical knowledge and experience may cause misdiagnoses or missed diagnoses in sore throat diagnoses, especially for general practitioners in primary hospitals. This study aims to develop a computer-aided diagnostic system to assist clinicians in the differential diagnoses of sore throat. The computer-aided system is developed based on the Dynamic Uncertain Causality Graph (DUCG) theory. We cooperated with medical specialists to establish a sore throat DUCG model as the diagnostic knowledge base. The construction of the model integrates epidemiological data, knowledge, and clinical experience of medical specialists. The chain reasoning algorithm of the DUCG is used for the differential diagnoses of sore throat. The system can diagnose 27 sore throat-related diseases. The model builder initially tests it with 81 cases, and all cases are correctly diagnosed. Then the system is verified by the third-party hospital, and the diagnostic accuracy is 98%. Now, the system has been applied in hundreds of primary hospitals in Jiaozhou City, China, and the degree of recognition for doctors to the diagnostic results of the system is more than 99.9%. It is feasible to use DUCG for the differential diagnoses of sore throat, which can assist primary doctors in clinical diagnoses and the diagnostic results are acceptable to clinicians.
... A single-assignment variable means that there is only one assignment for a variable, and a multiple-assignment variable means that there is more than one assignment for a variable. For detailed principles of DUCG theory, readers can refer to [58,[109][110][111][112]. ...
Fault diagnosis plays an important role in complex and safety-critical systems such as nuclear power plants (NPPs). With the development of artificial intelligence (AI), extensive research has been carried out for fast and efficient fault diagnosis based on intelligent methods. This paper presents a review of various AI-based system-level fault diagnosis methods for NPPs. We first discuss the development history of AI. Based on this exposition, AI-based fault diagnosis techniques are classified into knowledge-driven and data-driven approaches. For knowledge-driven methods, we discuss both the early if-then-based fault diagnosis techniques and the current new theory-based ones. The principles, application, and comparative analysis of the representative methods are systematically described. For data-driven strategies, we discuss single-algorithm-based techniques such as ANN, SVM, PCA, DT, and clustering, as well as hybrid techniques that combine algorithms together. The advantages and disadvantages of both knowledge-driven and data-driven methods are compared, illustrating the tendency to combine the two approaches. Finally, we provide some possible future research directions and suggestions.
... The DUCG can deal with causal loops, so the DUCG model supports the expression of causal loops [23]. The DUCG can deal with discrete, continuous, and fuzzy evidence, which increases the robustness of the model [24]. ...
Purpose
Epistaxis is a common symptom and can be caused by various diseases, including nasal diseases, systemic diseases, etc. Many misdiagnosis and missed diagnosis of epistaxis are caused by lack of clinical knowledge and experience, especially some interns and the clinicans in primary hospitals. To help inexperienced clinicans improve their diagnostic accuracies of epistaxis, a computer-aided diagnostic system based on Dynamic Uncertain Causality Graph (DUCG) was designed in this study.
Methods
We build a visual epistaxis knowledge base based on medical experts’ knowledge and experience. The knowledge base intuitively expresses the causal relationship among diseases, risk factors, symptoms, signs, laboratory checks, and image examinations. The DUCG inference algorithm well addresses the patients’ clinical information with the knowledge base to deduce the currently suspected diseases and calculate the probability of each suspected disease.
Result
The model can differentially diagnose 24 diseases with epistaxis as the chief complaint. A third-party verification was performed, and the total diagnostic precision was 97.81%. In addition, the DUCG-based diagnostic model was applied in Jiaozhou city and Zhongxian county, China, covering hundreds of primary hospitals and clinics. So far, the clinicians using the model have all agreed with the diagnostic results. The 432 real-world application cases show that this model is good for the differential diagnoses of epistaxis.
Conclusion
The results show that the DUCG-based epistaxis diagnosis model has high diagnostic accuracy. It can assist primary clinicians in completing the differential diagnosis of epistaxis and can be accepted by clinicians.
... The knowledge reasoning process of a DUCG can be divided into three steps [4][5][6][7]: (1) simplification of DUCG based on evidence to decrease its scale; (2) extending causality chain expression of the consequence events composing of independent events; and (3) calculating the probabilities of consequence events according to these expressions. In addition, a complicated DUCG can be represented by a set of uncomplicated sub-DUCGs in the construction of a knowledge base [8]. Due to its outstanding capability in depicting uncertain causalities and performing efficient reasoning, the DUCG has been widely used in many fields, such as medical diagnosis and treatment [9,10], fault diagnosis of nuclear power plants [11], reliability analysis of dynamic reliability block diagram [12], shale-gas sweet-spot evaluation [13], and probabilistic safety assessment of a boiling water reactor [14]. ...
A dynamic uncertain causality graph (DUCG) is a probabilistic graphical model for knowledge representation and reasoning, which has been widely used in many areas, such as probabilistic safety assessment, medical diagnosis, and fault diagnosis. However, the convention DUCG model fails to model experts’ knowledge precisely because knowledge parameters were crisp numbers or fuzzy numbers. In reality, domain experts tend to use linguistic terms to express their judgements due to professional limitations and information deficiency. To overcome the shortcomings of DUCGs, this article proposes a new type of DUCG model by integrating Pythagorean uncertain linguistic sets (PULSs) and the evaluation based on the distance from average solution (EDAS) method. In particular, experts express knowledge parameters in the form of the PULSs, which can depict the uncertainty and vagueness of expert knowledge. Furthermore, this model gathers the evaluations of experts on knowledge parameters and handles conflicting opinions among them. Moreover, a reasoning algorithm based on the EDAS method is proposed to improve the reliability and intelligence of expert systems. Lastly, an industrial example concerning the root cause analysis of abnormal aluminum electrolysis cell condition is provided to demonstrate the proposed DUCG model.
... Cause analysis is an important method to assess the priority of the abnormal events in a failure system. In recent decades, Bayesian network (BN) [1,2], dynamic uncertain causality graph (DUCG) [3,4], neural network (NN) [5,6] and fuzzy Petri net (FPN) [7][8][9] have been studied as powerful methods for fault analysis. These methods present the relationships between events using graphical symbols and inference rules, which can make the process of fault analysis more intuitive and clear. ...
Cause analysis makes great contributions to identifying the priorities of the causes in fault diagnosis system. A fuzzy Petri net (FPN) is a preferable model for knowledge representation and reasoning and has become an effective fault diagnosis tool. However, the existing FPN has some limitations in cause analysis. It is criticized for the inability to fully consider incomplete and unknown knowledge in uncertain situations. In this paper, an enhanced grey reasoning Petri net (EGRPN) based on matrix operations is presented to address the limitations and improves the flexibility of the existing FPN. The proposed EGRPN model uses grey numbers to handle the greyness and inaccuracy of uncertain knowledge. Then, the EGRPN inference algorithm is executed based on the matrix operations, which can express the relevance of uncertain events in the form of grey numbers and improve the reliability of the knowledge reasoning process. Finally, industrial examples of cause diagnosis are used to illustrate the feasibility and reliability of the EGRPN model. The experimental results show that the new EGRPN model is promising for cause analysis.
... The dynamic uncertain causality graph (DUCG) [1][2][3] is a significant graphical way for the establishment of knowledge-based systems and has received much attention by academic scholars in recent decades. The basic concepts of the DUCG are representation of causal relationships and probabilistic inference of uncertain events. ...
The dynamic uncertain causality graph (DUCG), which has been widely applied in many fields, is an important modelling technique for knowledge representation and reasoning. However, the extant DUCG models have been criticized because they cannot precisely represent experts’ knowledge owing to the ignorance of the fuzziness and randomness of uncertain knowledge. In response, we propose a new type of DUCG model called the cloud reasoning dynamic uncertain causality graph (CDUCG). The CDUCG model, which is based on cloud model theory, can handle with the fuzziness and randomness of uncertain information simultaneously. Moreover, an inference algorithm based on the combination of CDUCG and the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) is proposed to implement fuzzy knowledge inference effectively and thus make the expert systems more dependable and intelligent. Finally, illustrative examples and an industrial application concerning root cause analysis of aluminum electrolysis are provided to demonstrate the proposed CDUCG model. And experimental results show that the new CDUCG model is flexible and reliable for knowledge representation and reasoning.
... The dynamic uncertain causality graph (DUCG) is a newly presented approach to graphically and compactly represent complex uncertain causalities, and perform probabilistic reasoning [39]. DUCG is able to compactly and graphically represent uncertainty causality in many cases, simplify graphics, outspread events based on evidence observed, and calculate the updated probabilities of the queries still of concern [40]. Also, DUCG relies less on the accuracy of parameters and the integrity of knowledge expression. ...
To reduce the influence of uncertain factors on the results of gearbox operation condition evaluation and fault diagnosis, and to improve the reliability and stability of gearbox operation, an improved dynamic uncertain causality graph (DUCG) fault diagnosis method is proposed by combining the qualitative and quantitative information obtained. In addition, to address the lack of objectivity of correlation variables in the dynamic uncertainty causal graph, the combination weighting method is used to reassign correlation variables. The sub-DUCGs of gear, bearing, shaft and box are established and connected with logic gate and conditional connection variables. The DUCG is used to diagnose the faults in the gearbox, and the effectiveness and rationality of the method are verified by comparing the probabilities of the maximum pre-selected events before and after the improvement. Because the combination weighting method only makes moderate modifications for different weights, the limitations of the diagnosis accuracy and the calculation of variable weights are discussed by choosing faults with different numbers of weights. The results show that the improved DUCG can more accurately identify root faults, and the growth rate of the probability of maximum pre-selected event increases with an increase in the number of weights.
... R ELATION classification serves as a crucial step in knowledge extraction from raw unstructured texts, which plays a key role in various natural language processing (NLP) applications, such as information extraction (IE) [1], knowledge base completion (KBC) [2] and question answering (QA) [3]. For instance, along with the explosive growth of data, traditional rule-based QA systems can not satisfy the need of performance and knowledge quantity for users, therefore, various knowledge-based QA systems have been built and widely used in web search and other applications in recent years. ...
The goal of relation classification is to recognize the relationship between two marked entities in a sentence. It is a crucial constituent in natural language processing. Up till the present moment, most previous neural network models for this task either focus on using the handcrafted syntactic features or learning semantic representations of raw word sequences, they have no capacity for encoding whole sentence representation including syntax and semantic information. In general, information of syntax and semantics can both have significant effect on classifying relation. Based on this idea, we propose a novel two-channels neural network architecture with attention mechanism in the paper to handle this task. Firstly, we employ bidirectional sequence LSTM channel to capture the semantic information and acquire syntactic knowledge by utilizing tree structure LSTM channel. Secondly, sentence level attention mechanism for word sequences is used to determine which parts of the sentence are most influential component. Eventually, we conduct experiments on two real world datasets: the Wikipedia and the SemEval-2010 Task8 dataset. The experimental results on datasets demonstrate that our method can make better use of the information contained in sentences, and achieve impressive improvements on relation classification as compared with the existing methods.
... DUCG [2][3][4][5][6][7] is a probabilistic graphic model dealing with uncertain causality representation and probabilistic reasoning. DUCG consists of {B-type, X-type, D-type, G-type} nodes and {P-type, F-type} arcs, simple DUCG graphs are shown in Section 3-5. ...
Fault Tree Analysis (FTA) has been widely applied to large, complex industrial systems like nuclear power plants, chemical systems, and weapon systems. Events in classical FTA are assumed binary-state and s-independent but multi-state, dependencies and logic cycles may exist within FTs. Moreover, causalities in FTA are assumed deterministic, while sometimes they may be uncertain. This paper applies Dynamic Uncertain Causality Graph (DUCG) in FTA to overcome aforementioned issues. This paper shows that any FT can be mapped into a DUCG graph. And with DUCG representation model and algorithm, additional modeling and analytical power are obtained. Multi-value, dependencies, logic cycles, and non-deterministic causalities in FTA are solved. This paper also depicts how to calculate the importance measurement, predict failure, and diagnose fault. The results reveal the effectiveness and feasibility of this methodology.
... DUCG is a probabilistic graphical model which intuitively expresses a causal relationship among variables in an explicit pattern, and uses a "chaining" inference algorithm to achieve efficient reasoning. DUCG can propagate probabilities through causality chains, achieve dynamic reasoning either with or without spread of causality between time slices (Zhang and Geng, 2015), achieve reasoning in the case of logic circles (Zhang, 2015a), and handle fuzzy evidence (Zhang, 2015b). The greatest advantage of DUCG in clinical diagnosis is that it can display the reasoning process and results graphically, and make an inference with incomplete information and less accurate parameters than conventional methods such as Bayesian Networks. ...
Jaundice is a common and complex clinical symptom potentially occurring in hepatology, general surgery, pediatrics, infectious diseases, gynecology, and obstetrics, and it is fairly difficult to distinguish the cause of jaundice in clinical practice, especially for general practitioners in less developed regions. With collaboration between physicians and artificial intelligence engineers, a comprehensive knowledge base relevant to jaundice was created based on demographic information, symptoms, physical signs, laboratory tests, imaging diagnosis, medical histories, and risk factors. Then a diagnostic modeling and reasoning system using the dynamic uncertain causality graph was proposed. A modularized modeling scheme was presented to reduce the complexity of model construction, providing multiple perspectives and arbitrary granularity for disease causality representations. A “chaining” inference algorithm and weighted logic operation mechanism were employed to guarantee the exactness and efficiency of diagnostic reasoning under situations of incomplete and uncertain information. Moreover, the causal interactions among diseases and symptoms intuitively demonstrated the reasoning process in a graphical manner. Verification was performed using 203 randomly pooled clinical cases, and the accuracy was 99.01% and 84.73%, respectively, with or without laboratory tests in the model. The solutions were more explicable and convincing than common methods such as Bayesian Networks, further increasing the objectivity of clinical decision-making. The promising results indicated that our model could be potentially used in intelligent diagnosis and help decrease public health expenditure.
... Алгебраические байесовские сети (АБС) являются одним из представлений баз знаний с неопределенностью [1][2][3]. Они родственны байесовским сетям доверия, широко используемым в оценке рисков [4,5], системах поддержки принятия решений [6] и прогнозировании [7]. Классическим представлением АБС является граф, в вершинах которого стоят фрагменты знаний, являющиеся совокупностями достаточно тесно связанных объектов, которым приписаны оценки вероятностей, а ребра представляют логические связи между вершинами, обладающими общими элементами [3,8]. ...
Subject of Research. Algebraic Bayesian networks are referred to a class of probabilistic graphical models that are a representation of knowledge bases with uncertainty. The distinguishing feature of ABN is the availability of global structures. Among them there are primary and secondary structures that are directly used in various kinds of probabilistic logical inference as well as tertiary and quaternary involved in the problems of automatic synthesis and identification of the properties of the secondary structure and partially in the machine learning tasks within specified networks. Existing algorithms for quaternary structure changing require its complete rebuild when changing the primary structure. That feature slows down the whole global structures synthesis, dispels user’s attention who is forced to re-analyze the whole rebuilt structure instead of focusing on the changes that were directly caused by the limited modification of the original data. This fact reduces ABN attractiveness as a model for data processing in general. Scope of Research. This paper is aimed at speeding up the rebuild process and eliminating the shortage of the quaternary structure rebuild algorithms when adding and deleting vertices of primary structure expressed in excess rebuild of the entire structure. The task of algorithm incrementalization for quaternary structure rebuild is solved to achieve the goal. Method. The proposed approach is based on the properties of incremental algorithms that reduce the amount of computations due to the result obtained at the previous step of the algorithm. All the arguments used in the paper are expressed in a graph theory language to apply the established system of terms and classical results. Main Results. The paper presents incremental and decremental algorithms, complemented by a proof of correctness and listing. Given algorithms are based on the previously obtained incremental algorithms for tertiary structure. Moreover, a detailed analysis of the plurality of separators is carried out on each stage of the algorithms. Theoretical and Practical Relevance. These algorithms develop a global structure of algebraic Bayesian networks as well as the theory of probabilistic graphical models in general. Furthermore, they create the groundwork for creation of the secondary structure invariants that may be non-unique even if the primary structure is fixed unlike the current approach where connections are included in the secondary structure. The deletion of manipulation with a set of secondary structures considerably simplifies and makes foreseeable visualization of such complex object as ABN and may improve the computational characteristics of probabilistic logical inference algorithms. It also enables to reformulate the problem of ABN machine learning excluding any need for synthesis of many different objects with complex structure and virtually the same semantics. It also can be expected that the obtained incremental algorithms will accelerate computational processes of rebuilding and properties analysis for all four global ABN structures.
AI-aided clinical diagnosis is desired in medical care. Existing deep learning models lack explainability and mainly focus on image analysis. The recently developed Dynamic Uncertain Causality Graph (DUCG) approach is causality-driven, explainable, and invariant across different application scenarios, without problems of data collection, labeling, fitting, privacy, bias, generalization, high cost and high energy consumption. Through close collaboration between clinical experts and DUCG technicians, 46 DUCG models covering 54 chief complaints were constructed. Over 1,000 diseases can be diagnosed without triage. Before being applied in real-world, the 46 DUCG models were retrospectively verified by third-party hospitals. The verified diagnostic precisions were no less than 95%, in which the diagnostic precision for every disease including uncommon ones was no less than 80%. After verifications, the 46 DUCG models were applied in the real-world in China. Over one million real diagnosis cases have been performed, with only 17 incorrect diagnoses identified. Due to DUCG's transparency, the mistakes causing the incorrect diagnoses were found and corrected. The diagnostic abilities of the clinicians who applied DUCG frequently were improved significantly. Following the introduction to the earlier presented DUCG methodology, the recommendation algorithm for potential medical checks is presented and the key idea of DUCG is extracted.
AI-aided clinical diagnosis is desired in medical care. Existing deep learning models lack explainability and mainly focus on image analysis. The recently developed Dynamic Uncertain Causality Graph (DUCG) approach is causality-driven, explainable, and invariant across different application scenarios, without problems of data collection, labeling, fitting, privacy, bias, generalization, high cost and high energy consumption. Through close collaboration between clinical experts and DUCG technicians, 46 DUCG models covering 54 chief complaints were constructed. Over 1,000 diseases can be diagnosed without triage. Before being applied in real-world, the 46 DUCG models were retrospectively verified by third-party hospitals. The verified diagnostic precisions were no less than 95%, in which the diagnostic precision for every disease including uncommon ones was no less than 80%. After verifications, the 46 DUCG models were applied in the real-world in China. Over one million real diagnosis cases have been performed, with only 17 incorrect diagnoses identified. Due to DUCG’s transparency, the mistakes causing the incorrect diagnoses were found and corrected. The diagnostic abilities of the clinicians who applied DUCG frequently were improved significantly. Following the introduction to the earlier presented DUCG methodology, the recommendation algorithm for potential medical checks is presented and the key idea of DUCG is extracted.
Root cause diagnosis can locate abnormalities of industrial processes, ensuring production safety and manufacturing efficiency. However, existing root cause diagnosis models only consider pairwise direct causality and ignore the multi-level fault propagation, which may lead to incomplete root cause descriptions and ambiguous root cause candidates. To address the above issue, a novel framework, named multi-level predictive graph extraction (MPGE) and RootRank scoring, is proposed and applied to the root cause diagnosis for industrial processes. In this framework, both direct and indirect Granger causalities are characterized by multi-level predictive relationships to provide a sufficient characterization of root cause variables. First, a predictive graph structure with a sparse constrained adjacency matrix is constructed to describe the information transmission between variables. The information of variables is deeply fused according to the adjacency matrix to consider multi-level fault propagation. Then, a hierarchical adjacency pruning (HAP) mechanism is designed to automatically capture vital predictive relationships through adjacency redistribution. In this way, the multi-level causalities between variables are extracted to fully describe both direct and indirect fault propagation and highlight the root cause. Further, a RootRank scoring algorithm is proposed to analyze the predictive graph and quantify the fault propagation contribution of each variable, thereby giving definite root cause identification results. Three examples are adopted to verify the diagnostic performance of the proposed framework, including a numerical example, the Tennessee Eastman benchmark process, and a real cut-made process of cigarette. Both theoretical analysis and experimental verification show the high interpretability and reliability of the proposed framework.
Representing and reasoning uncertain causalities have diverse applications in fault diagnosis for industrial systems. Owing to the complicated dynamics and a multitude of uncertain factors in such systems, it is hard to implement efficient diagnostic inference when a process fault occurs. The cubic dynamic uncertain causal graph was proposed for graphically modeling and reasoning about the fault spreading behaviors in the form of causal dependencies across multivariate time series. However, in certain large-scale scenarios with multiconnected and time-varying causalities, the existing inference algorithm is incapable of dealing with the logical reasoning process in an efficient manner. We, therefore, explore the solutions to enhanced computational efficiency. Causality graph decomposition and simplification, and graphical transformation, are proposed to reduce model complexity and form a minimal causality graph. An algorithm, event-oriented early logical absorption, is also presented for logical reasoning. It is mainly intended to minimize the computational costs for compulsory absorption operations in the early stage of reasoning process. The effectiveness of the proposed algorithm is verified on the secondary loop model of nuclear power plant by utilizing the fault data derived from a nuclear power plant simulator. The results show the anticipated capability of efficient fault diagnosis of the proposed algorithm for large-scale dynamic systems.
This study presents an industrial fault diagnosis system based on the cubic dynamic uncertain causality graph (cubic DUCG) used to model and diagnose industrial systems without sufficient data for model training. The system is developed based on cloud native technology. It contains two main parts, the diagnostic knowledge base and the inference method. The knowledge base was built by domain experts modularly based on professional knowledge. It represented the causality between events in the target industrial system in a visual and graphical form. During the inference, the cubic DUCG algorithm could dynamically generate the cubic causal graph according to the real-time data and perform the logic and probability calculations based on the generated cubic DUCG models, visually displaying the dynamic causal evolution of faults. To verify the system’s feasibility, we rebuild a fault-diagnosis model of the secondary circuit system of No. 1 at the Ningde nuclear power plant based on the new system. Twenty-four fault cases were used to test the diagnostic accuracy of the system, and all faults were correctly diagnosed. The results showed that it was feasible to use the cubic DUCG platform for fault diagnosis.
Artificial intelligence (AI)-aided general clinical diagnosis is helpful to primary clinicians. Machine learning approaches have problems of generalization, interpretability, etc. Dynamic Uncertain Causality Graph (DUCG) based on uncertain casual knowledge provided by clinical experts does not have these problems. This paper extends DUCG to include the representation and inference algorithm for non-causal classification relationships. As a part of general clinical diagnoses, six knowledge bases corresponding to six chief complaints (arthralgia, dyspnea, cough and expectoration, epistaxis, fever with rash and abdominal pain) were constructed through constructing subgraphs relevant to a chief complaint separately and synthesizing them together as the knowledge base of the chief complaint. A subgraph represents variables and causalities related to a single disease that may cause the chief complaint, regardless of which hospital department the disease belongs to. Verified by two groups of third-party hospitals independently, total diagnostic precisions of the six knowledge bases ranged in 96.5–100%, in which the precision for every disease was no less than 80%.
Root cause diagnosis is of great significance to make efficient decisions in industrial production processes. It is a procedure of fusing knowledge, such as empirical knowledge, process knowledge, and mechanism knowledge. However, it is insufficient and low reliability of cause analysis methods by using crisp values or fuzzy numbers to represent uncertain knowledge. Therefore, a dynamic uncertain causality graph model (DUCG) based on picture fuzzy set (PFS) is proposed to address the problem of uncertain knowledge representation and reasoning. It combines the PFS with DUCG model to express expert doubtful ideas in a complex system. Then, a new PFS operator is introduced to characterize the importance of factors and connections among various information. Moreover, an enhanced knowledge reasoning algorithm is developed based on the PFS operators to resolve causal inference problems. Finally, a numerical example illustrates the effectiveness of the method, and the results show that the proposed model is more reliable and flexible than the existing models.
Modeling a complex system with a large number of variables at one time and collecting global data on a large scale is usually not practical. It is ideal to model the global system according to only local data and structures, and then synthesize them together. Multiple local structures can be easily synthesized as a global structure through fusing same nodes, while multiple sets of local data cannot be combined as the global data. Therefore, we need a method to obtain the joint probability distribution (JPD) of the global system by utilizing local data only. When the global structure synthesized does not include directed cyclic graph (DCG), it is easy to calculate the JPD with Bayesian network (BN). When DCGs are included, BN does not work. This paper presents the multi-valued Dynamic Uncertain Causality Graph (M-DUCG) methodology to calculate the global JPD in the case of DCGs with only local data. The idea is to use multiple sets of local data to learn the parameters of the synthesized M-DUCG structure with DCGs, and then use these parameters to calculate the global JPD.
Many AI systems have been developed for clinical diagnoses, in which most of them lack interpretability in both knowledge representation and inference results. The newly developed Dynamic Uncertain Causality Graph (DUCG) is a probabilistic graphical model with strong interpretability. However, existing DUCG is mainly for fault diagnoses of large, complex industrial systems. In this paper, we extend DUCG for better application in general clinical diagnoses. Four extensions are introduced: (1) special logic gate and zoom function event variables to represent and quantify the influences of various risk factors on the morbidities of diseases. (2) Reversal logic gate to model the case that some diseases/causes may result in at least two simultaneous symptoms/consequences. (3) Disease-specific manifestation variable for special inference and easy understanding to diagnose a specific disease. (4) Event attention importance to count contributions of isolated state-abnormal variables in inference. To illustrate and verify the extended DUCG methodology, we performed a case study for diagnosing 25 diseases causing nasal obstruction. We tested 171 cases randomly selected from total 471 cases of discharged patients in the hospital information system of Xuanwu Hospital. The diagnosis precision of the extended DUCG was 100%. The diagnosis precision of the third-party verification performed by Suining Central Hospital was 98.86%, which exhibited the strong generalization ability of the extended DUCG.
The energy-intensive enterprises (EIEs), such as iron and steel enterprises, account for a significant part of the total energy consumption in society. The Linz-Donawitz converter gas (LDG) is a kind of crucial byproduct energy resource recycled during the steelmaking process, and its reasonable scheduling can effectively reduce the LDG emission and increase its efficiency. In this study, a granular-causality-based scheduling approach for the LDG system in steel industry is proposed. A granular causality technique is modeled to confirm the casual relationship of the LDG system based on the discontinuous production characteristic, in which a causality diagram is established and the phase space of the training sample is reconstructed to improve the prediction accuracy. Then, a multioutput least-square support vector machine model is constructed for the prediction of the gas tank levels. In order to consider the impact of multiple solutions on the scheduling result in a period of time, a scheduling objective function that combines the economy criterion and the safety one is designed and optimized by a modified particle swarm optimization (PSO) algorithm. The validation experiments using real-world data coming from the energy data center of a steel plant are carried out, and the results indicate that the proposed method exhibits reliable performance. Moreover, an application software system based on the proposed method is developed and implemented, which demonstrates the applicability of the proposed approach.
In the real world, an attribute value is perceived differently by different individuals. Emphasizing on this aspect, we extend the discrete choice models with perceived values that are subjective and specific to a decision-maker (DM). The proposed choice models are augmented with the parameters of an entropy function, besides the utility coefficients, to model a DM's complex choice behavior. A variety of higher order choice models are also proposed. The proposed models are further extended with a DM's reference-value for each attribute. A real and illustrative application is included.
Scientific modeling and analysis for fault spreading process is a promising way for guaranteeing the safe, reliable, and efficient operation of complex system. However, the representing and reasoning of uncertain, time-varying, and sophisticated dependences are difficult, especially for the complex issues of dynamic negative feedback loops in multivariate time series. Dynamic uncertain causality graph (DUCG) provides a dynamic inference method without causality propagation across time slices, the disadvantage of which lies in the interpretability and applicability. In order to overcome the shortcomings of DUCG and extend its capabilities of temporal causality representation and dynamic reasoning, this paper proposes a new methodology named Cubic DUCG. The fundamental idea is to continuously generate the cubic causality graph online according to the sequential observations by discarding the restrictive Markov and conditional independence assumptions. Based on the complete causal dependencies representing the real-time fault spreading behaviors, the efficient and rigorous inference algorithm is thus proposed. The method is validated on fault data of the secondary loop of two nuclear power plant simulators, concerning the effectiveness, and in particular, the capability of dealing with complex dynamics and negative feedback processes.
When the process is working normally, process operating performance optimality assessment makes a further judgment on how well the process is running currently. Sustained optimal process operating performance guarantees the greatest economic benefits. However, the operating performance may deteriorate from the optimal status due to the production condition shifts. It is crucial to accomplish operating performance optimality assessment strategies for a deeper understanding of the process status. A Dynamic Causality Diagram (DCD) explicitly and intuitively describes the causal correlations between variables. However, an actual process has both quantitative and qualitative measurements, which makes the conventional DCD imprecise or even invalid. To handle the above problems, a modified DCD is proposed to assess a complicated process with coexistence of quantitative and qualitative information. An operating performance optimality assessing strategy based on the proposed DCD is then established. The proposed method is successfully applied to a gold hydrometallurgy process, and the operating performance optimality assessment result is satisfied. This article is protected by copyright. All rights reserved
The dynamic uncertain causality graph (DUCG) is a newly presented framework for uncertain causality representation and probabilistic reasoning. It has been successfully applied to online fault diagnoses of large, complex industrial systems, and decease diagnoses. This paper extends the DUCG to model more complex cases than what could be previously modeled, e.g., the case in which statistical data are in different groups with or without overlap, and some domain knowledge and actions (new variables with uncertain causalities) are introduced. In other words, this paper proposes to use i-mode, e-mode, and h-mode of the DUCG to model such complex cases and then transform them into either the standard i-mode or the standard e-mode. In the former situation, if no directed cyclic graph is involved, the transformed result is simply a Bayesian network (BN), and existing inference methods for BNs can be applied. In the latter situation, an inference method based on the DUCG is proposed. Examples are provided to illustrate the methodology.
Intelligent diagnosis system is applied to fault diagnosis in spacecraft. Dynamic Uncertain Causality Graph (DUCG) is a new probability graphic model with many advantages. In this paper, DUGG is applied to fault diagnosis in spacecraft: introducing conditional functional events into ordinary DUCG to deal with spacecraft multi-conditions. Now, DUCG has been tested in 16 typical faults with 100% diagnosis accuracy.
A significant amount of research has been devoted to seeking efficient solutions to the problem of pattern matching over graphs. This interest is largely due to the many applications that require such efficient solutions, including protein complex prediction, social network analysis, and structural pattern recognition. However, in many real applications, the graph data are often noisy, incomplete, and inaccurate. In other words, there exist many uncertain graphs. Therefore, in this paper, we study pattern matching in the context of large uncertain graphs. Specifically, we want to retrieve all qualified matches of a query pattern in the uncertain graph. Though pattern matching over uncertain graphs is NP-hard, we employ a filtering-and-verification framework to speed up the search. In the filtering phase, we propose a probabilistic matching tree (PM-tree) built from match cuts obtained by a cut selection process. Based on the PM-tree, we devise a collective pruning strategy to prune a large number of unqualified matches. During the verification phase, we develop an efficient sampling algorithm to validate the remaining candidates. Extensive experimental results demonstrate the effectiveness and efficiency of the proposed algorithms. Finally, we show how our solution can be applied to querying knowledge graphs.
Intelligent systems are desired in dynamic fault diagnoses for large and complex systems such as nuclear power plants. Dynamic uncertain causality graph (DUCG) is such a system presented previously. This paper extends the DUCG methodology to deal with negative feedbacks, which is one of the most difficult problems in fault diagnosis, and predicts the fault development online. Two methods are presented. One does not involve causality propagation across time slices. Another involves causality propagation through time slices. A nuclear power plant simulator located at Tsinghua University is used to test the DUCG methodology. A typical experiment involving negative feedback is given, which jointly applies the methods separately presented in different papers, such as how to deal with dynamic fault diagnosis and directed cyclic graphs (DCGs), and the methods presented in this paper. Results show that DUCG is powerful in knowledge representation, diagnosing possible faults, and predicting developments of faults. It also demonstrates that DUCG is robust, which means that the DUCG inference does not rely much on the accuracy of probability parameters. The logics of diagnoses and predictions are graphically displayed, so that users know not only inference results, but also why they are correct.
Probabilistic graphical models (PGMs) such as Bayesian network (BN) have been widely applied in uncertain causality representation and probabilistic reasoning. Dynamic uncertain causality graph (DUCG) is a newly presented model of PGMs, which can be applied to fault diagnosis of large and complex industrial systems, disease diagnosis, and so on. The basic methodology of DUCG has been previously presented, in which only the directed acyclic graph (DAG) was addressed. However, the mathematical meaning of DUCG was not discussed. In this paper, the DUCG with directed cyclic graphs (DCGs) is addressed. In contrast, BN does not allow DCGs, as otherwise the conditional independence will not be satisfied. The inference algorithm for the DUCG with DCGs is presented, which not only extends the capabilities of DUCG from DAGs to DCGs but also enables users to decompose a large and complex DUCG into a set of small, simple sub-DUCGs, so that a large and complex knowledge base can be easily constructed, understood, and maintained. The basic mathematical definition of a complete DUCG with or without DCGs is proved to be a joint probability distribution (JPD) over a set of random variables. The incomplete DUCG as a part of a complete DUCG may represent a part of JPD. Examples are provided to illustrate the methodology.
Graphical models for probabilistic reasoning are now in widespread use. Many approaches have been developed such as Bayesian network. A newly developed approach named as dynamic uncertain causality graph (DUCG) is initially presented in a previous paper, in which only the inference algorithm in terms of individual events and probabilities is addressed. In this paper, we first explain the statistic basis of DUCG. Then, we extend the algorithm to the form of matrices of events and probabilities. It is revealed that the representation of DUCG can be incomplete and the exact probabilistic inference may still be made. A real application of DUCG for fault diagnoses of a generator system of a nuclear power plant is demonstrated, which involves variables. Most inferences take with a laptop computer. The causal logic between inference result and observations is graphically displayed to users so that they know not only the result, but also why the result obtained.
The Bayesian Networks are graphical models that are easy to interpret and update. These models are useful if the knowledge is uncertain, but they lack some means to express ambiguity. To face this problem, we propose Fuzzy Evidence in Bayesian Networks and combine the Fuzzy Logic and Bayesian Network. This has allowed to benefit from mutual advantages of these two approaches, and to overcome the problem of data and observation ambiguity. This paper proposes an inference algorithm which uses the Bayesian Network and Fuzzy Logic reliability. This solution has been implemented, tested and evaluated in comparison with the existing methods.
The Bayesian Networks are graphical models that are easy to interpret and update. These models are useful if the knowledge is uncertain, but they lack some means to express ambiguity. To face this problem, we propose Fuzzy Evidence in Bayesian Networks and combine the Fuzzy Logic and Bayesian Network. This has allowed to benefit from mutual advantages of these two approaches, and to overcome the problem of data and observation ambiguity. This paper proposes an inference algorithm which uses the Bayesian Network and Fuzzy Logic reliability. This solution has been implemented, tested and evaluated in comparison with the existing methods.
We show how to use a variational approximation to the logistic function to
perform approximate inference in Bayesian networks containing discrete nodes
with continuous parents. Essentially, we convert the logistic function to a
Gaussian, which facilitates exact inference, and then iteratively adjust the
variational parameters to improve the quality of the approximation. We
demonstrate experimentally that this approximation is faster and potentially
more accurate than sampling. We also introduce a simple new technique for
handling evidence, which allows us to handle arbitrary distributions on
observed nodes, as well as achieving a significant speedup in networks with
discrete variables of large cardinality.
Many real life domains contain a mixture of discrete and continuous variables
and can be modeled as hybrid Bayesian Networks. Animportant subclass of hybrid
BNs are conditional linear Gaussian (CLG) networks, where the conditional
distribution of the continuous variables given an assignment to the discrete
variables is a multivariate Gaussian. Lauritzen's extension to the clique tree
algorithm can be used for exact inference in CLG networks. However, many
domains also include discrete variables that depend on continuous ones, and CLG
networks do not allow such dependencies to berepresented. No exact inference
algorithm has been proposed for these enhanced CLG networks. In this paper, we
generalize Lauritzen's algorithm, providing the first "exact" inference
algorithm for augmented CLG networks - networks where continuous nodes are
conditional linear Gaussians but that also allow discrete children ofcontinuous
parents. Our algorithm is exact in the sense that it computes the exact
distributions over the discrete nodes, and the exact first and second moments
of the continuous ones, up to the accuracy obtained by numerical integration
used within thealgorithm. When the discrete children are modeled with softmax
CPDs (as is the case in many real world domains) the approximation of the
continuous distributions using the first two moments is particularly accurate.
Our algorithm is simple to implement and often comparable in its complexity to
Lauritzen's algorithm. We show empirically that it achieves substantially
higher accuracy than previous approximate algorithms.
An important class of continuous Bayesian networks are those that have linear conditionally deterministic variables (a variable that is a linear deterministic function of its parents). In this case, the joint density function for the variables in the network does not exist. Conditional linear Gaussian (CLG) distributions can handle such cases when all variables are normally distributed. In this paper, we develop operations required for performing inference with linear conditionally deterministic variables in continuous Bayesian networks using relationships derived from joint cumulative distribution functions. These methods allow inference in networks with linear deterministic variables and non-Gaussian distributions.
This paper reports our investigation on the problem of belief update in Bayesian networks (BN) using uncertain evidence. We focus on two types of uncertain evidences, virtual evidence (represented as likelihood ratios) and soft evidence (represented as probability distributions). We review three existing belief update methods with un- certain evidences: virtual evidence method, Jeffrey's rule, and IPFP (iterative proportional fitting procedure), and analyze the relations between these methods. This in- depth understanding leads us to propose two algorithms for belief update with multiple soft evidences. Both of these algorithms can be seen as integrating the tech- niques of virtual evidence method, IPFP and traditional BN evidential inference, and they have clear computa- tional and practical advantages over the methods pro- posed by others in the past.
We present in this article a case study of the prob- abilistic approach to model-based diagnosis. Here, the diagnosed system is a real-world electrical power system, namely the Advanced Diagnostic and Prognostic Testbed (ADAPT) located at the NASA Ames Research Center. Our probabilistic approach is formally well-founded, and based on Bayesian networks and arithmetic circuits. We pay special attention to meeting two of the main challenges — model development and real-time reasoning — often associated with real-world application of model-based diagnosis technologies. To address the challenge of model develop- ment, we develop a systematic approach to representing electrical power systems as Bayesian networks, supported by an easy-to- use specication language. To address the real-time reasoning challenge, we compile Bayesian networks into arithmetic circuits. Arithmetic circuit evaluation supports real-time diagnosis by being predictable and fast. In experiments with the ADAPT Bayesian network, which contains 503 discrete nodes and 579 edges and produces accurate results, the time taken to compute the most probable explanation using arithmetic circuits has a mean of 0.2625 milliseconds and a standard deviation of 0.2028 milliseconds. In comparative experiments, we found that while the variable elimination and join tree propagation algorithms also perform very well in the ADAPT setting, arithmetic circuit evaluation was an order of magnitude or more faster.
Computer vision techniques have made considerable progress in recognizing object categories by learning models that normally rely on a set of discriminative features. However, in contrast to human perception that makes extensive use of logic-based rules, these models fail to benefit from knowledge that is explicitly provided. In this paper, we propose a framework that can perform knowledge-assisted analysis of visual content. We use ontologies to model the domain knowledge and a set of conditional probabilities to model the application context. Then, a Bayesian network is used for integrating statistical and explicit knowledge and performing hypothesis testing using evidence-driven probabilistic inference. In addition, we propose the use of a focus-of-attention (FoA) mechanism that is based on the mutual information between concepts. This mechanism selects the most prominent hypotheses to be verified/tested by the BN, hence removing the need to exhaustively test all possible combinations of the hypotheses set. We experimentally evaluate our framework using content from three domains and for the following three tasks: 1) image categorization; 2) localized region labeling; and 3) weak annotation of video shot keyframes. The results obtained demonstrate the improvement in performance compared to a set of baseline concept classifiers that are not aware of any context or domain knowledge. Finally, we also demonstrate the ability of the proposed FoA mechanism to significantly reduce the computational cost of visual inference while obtaining results comparable to the exhaustive case.
A Bayesian network is a probabilistic representation for uncertain
relationships, which has proven to be useful for modeling real-world
problems. When there are many potential causes of a given effect,
however, both probability assessment and inference using a Bayesian
network can be difficult. In this paper, we describe causal
independence, a collection of conditional independence assertions and
functional relationships that are often appropriate to apply to the
representation of the uncertain interactions between causes and effect.
We show how the use of causal independence in a Bayesian network can
greatly simplify probability assessment as well as probabilistic
inference
Updated and expanded, Bayesian Artificial Intelligence, Second Edition provides a practical and accessible introduction to the main concepts, foundation, and applications of Bayesian networks. It focuses on both the causal discovery of networks and Bayesian inference procedures. Adopting a causal interpretation of Bayesian networks, the authors discuss the use of Bayesian networks for causal modeling. They also draw on their own applied research to illustrate various applications of the technology. New to the Second Edition New chapter on Bayesian network classifiers New section on object-oriented Bayesian networks New section that addresses foundational problems with causal discovery and Markov blanket discovery New section that covers methods of evaluating causal discovery programs Discussions of many common modeling errors New applications and case studies More coverage on the uses of causal interventions to understand and reason with causal Bayesian networks Illustrated with real case studies, the second edition of this bestseller continues to cover the groundwork of Bayesian networks. It presents the elements of Bayesian network technology, automated causal discovery, and learning probabilities from data and shows how to employ these technologies to develop probabilistic expert systems. Web ResourceThe books website at www.csse.monash.edu.au/bai/book/book.html offers a variety of supplemental materials, including example Bayesian networks and data sets. Instructors can email the authors for sample solutions to many of the problems in the text.
This paper aims to get a better understanding of the notions of evidence, probabilistic evidence and likelihood evidence in Bayesian Networks. Evidence comes from an observation of one or several variables. Soft evidence is probabilistic evidence, since the observation consists in a local probability distribution on a subset of variables that has to replace any former belief on these variables. It has to be clearly distinguished from likelihood evidence, also called virtual evidence, for which the evidence is specified as a likelihood ratio. Since the notion of soft evidence is not yet widely understood, most of the Bayesian Networks engines do not propose related propagation functions and the terms used to describe such evidence are not stabilised. First, we present the different types of evidence on a simple example with an illustrative context. Then, we discuss the understanding of both notions in terms of knowledge and observation. Next, we propose to use soft evidence to represent certain evidence on a continuous variable, after fuzzy discretization.
One problem faced in knowledge engineering for Bayesian networks (BNs) is the exponential growth of the number of parameters in their conditional probability tables (CPTs). The most common practical solution is the application of the so-called canonical gates and, among them, the noisy-OR (or their generalization, the noisy-MAX) gates, which take advantage of the independence of causal interactions and provide a logarithmic reduction of the number of parameters required to specify a CPT.
In this paper, we propose an algorithm that fits a noisy-MAX distribution to an existing CPT, and we apply this algorithm to search for noisy-MAX gates in three existing practical BN models: ALARM, HAILFINDER, and HEPAR II. We show that the noisy-MAX gate provides a surprisingly good fit for as many as 50% of CPTs in two of these networks. We observed this in both distributions elicited from experts and those learned from data. The importance of this finding is that it provides an empirical justification for the use of the noisy-MAX gate as a powerful knowledge engineering tool.
Vertigo is a common complaint with many potential causes involving otology, neurology and general medicine, and it is fairly difficult to distinguish the vertiginous disorders from each other accurately even for experienced physicians. Based on comprehensive investigations to relevant characteristics of vertigo, we propose a diagnostic modeling and reasoning methodology using Dynamic Uncertain Causality Graph. The symptoms, signs, findings of examinations, medical histories, etiology and pathogenesis, and so on, are incorporated in the diagnostic model. A modularized modeling scheme is presented to reduce the difficulty in model construction, providing multiple perspectives and arbitrary granularity for disease causality representations. We resort to the "chaining" inference algorithm and weighted logic operation mechanism, which guarantee the exactness and efficiency of diagnostic reasoning under situations of incomplete and uncertain information. Moreover, the causal insights into underlying interactions among diseases and symptoms intuitively demonstrate the reasoning process in a graphical manner. These solutions make the conclusions and advices more explicable and convincing, further increasing the objectivity of clinical decision-making. Verification experiments and empirical evaluations are performed with clinical vertigo cases. The results reveal that, even with incomplete observations, this methodology achieves encouraging diagnostic accuracy and effectiveness. This study provides a promising assistance tool for physicians in diagnosis of vertigo.
This paper demonstrates the use of graphs as a mathematical tool for expressing independencies, and as a formal language for communicating and processing causal information for decision analysis. We show how complex information about external interventions can be organized and represented graphically and, conversely, how the graphical representation can be used to facilitate quantitative predictions of the effects of interventions.
We first review the theory of Bayesian networks and show that directed acyclic graphs (DAGs) offer an economical scheme for representing conditional independence assumptions and for deducing and displaying all the logical consequences of such assumptions. We then introduce the manipulative account of causation and show that any DAG defines a simple transformation which tells us how the probability distribution will change as a result of external interventions in the system. Using this transformation it is possible to quantify, from non-experimental data, the effects of external interventions and to specify conditions under which randomized experiments are not necessary. As an example, we show how the effect of smoking on lung cancer can be quantified from non-experimental data, using a minimal set of qualitative assumptions.
Finally, the paper offers a graphical interpretation for Rubin’s model of causal effects, and demonstrates its equivalence to the manipulative account of causation. We exemplify the tradeoffs between the two approaches by deriving nonparametric bounds on treatment effects under conditions of imperfect compliance.
The authors have constructed an iterative, probabilistic reasoning
architecture for classification problems. A number of assumptions of
conditional independence have been employed in this architecture to
derive two iterative updating methods, S and D. A Bayesian network was
constructed and the results compared with the iterative methods. Method
S and the network are both insensitive to the order of evidence, but do
not produce the same results. Further investigation of the nature of
these differences is warranted. It is suggested that additional
information carried in the network may allow uncertain evidence to be
used more effectively than in the iterative methods
Conditional probability is extended so as to be conditioned by an uncertain proposition of which truth value is ¿ (0 ¿ ¿ ¿ 1). Using the extended conditional probability, the measure of increased belief and disbelief in a hypothesis resulting from the observation of uncertain evidence are derived from MYCIN's measures based on certain evidence. On the comparison with our measures, it is shown that MYCIN's intuitive measures based on uncertain evidence, called the strength of evidence, utilize affirmative information which increases belief in the uncertain evidence but ignore negative information which causes new doubt in the uncertain evidence. An interpretation of the disregard of the negative information is presented from the viewpoint of cognitive psychology. It is pointed out that this disregard of the negative information is reasonable for a model of human inference but the negative information must also be utilized in order to evaluate a hypothesis correctly, or impartially on the basis of uncertain evidence. Our measures provide a means for utilizing both the affirmative and negative information on uncertain evidence. It is shown that inference based on the negation of evidence, which is contained in one of our measures, is difficult for an expert. A method for estimating the measure is presented which does not demand the difficult inference from an expert. The significance of the method is explained from the viewpoint of cognitive psychology.
We revisit the problem of revising probabilistic beliefs using uncertain evidence, and report results on several major issues relating to this problem: how should one specify uncertain evidence? How should one revise a probability distribution? How should one interpret informal evidential statements? Should, and do, iterated belief revisions commute? And what guarantees can be offered on the amount of belief change induced by a particular revision? Our discussion is focused on two main methods for probabilistic revision: Jeffrey's rule of probability kinematics and Pearl's method of virtual evidence, where we analyze and unify these methods from the perspective of the questions posed above.
Developed from the dynamic causality diagram (DCD) model, a new approach for knowledge representation and reasoning named as dynamic uncertain causality graph (DUCG) is presented, which focuses on the compact representation of complex uncertain causalities and efficient probabilistic inference. It is pointed out that the existing models of compact representation and inference in Bayesian Network (BN) is applicable in single-valued cases, but may not be suitable to be applied in multi-valued cases. DUCG overcomes this problem and beyond. The main features of DUCG are: 1) compactly and graphically representing complex conditional probability distributions (CPDs), regardless of whether the cases are single-valued or multi-valued; 2) able to perform exact reasoning in the case of the incomplete knowledge representation; 3) simplifying the graphical knowledge base conditional on observations before other calculations, so that the scale and complexity of problem can be reduced exponentially; 4) the efficient two-step inference algorithm consisting of (a) logic operation to find all possible hypotheses in concern for given observations and (b) the probability calculation for these hypotheses; and 5) much less relying on the parameter accuracy. An alarm system example is provided to illustrate the DUCG methodology.
Two approaches for extending a quantitative Bayesian network (BN) to deal with qualitative information include a qualitative probability network extension and causal independence. Both approaches help developers to remedy the gap between the complicated BN formalism and the actual problem. Lucas utilizes these two methods in establishing a theory of qualitative causal (QC) interaction patterns where qualitative probability influences (QPIs) (considering whether it is better to hold a single cause than not) affect the network model. QC patterns help to offer developers a high-level starting point when developing BNs. However, in real-world applications, usually, we need to know QPIs on multiple causes, namely, whether holding some subset of the causes will be a better choice than holding other subsets. To this end, we introduce a concept called causality probability from which QPIs can be easily induced. We investigate the local optima and global optima of causality probabilities considering multiple causes. For the local optima, we present the qualitative influences on all binary interaction types, while for the global optima, we achieve an upper bound of causality probability and discuss the conditions to reach the upper bound. Our results are useful for BN developers to get an overview of causality relations.
A reasoning system and associated assessment methodology built on
a natural schema for an evidential argument are discussed. This argument
schema is based on the underlying causal chains linking conclusions and
evidence. The framework couples a probabilistic calculus with
qualitative approaches to evidential reasoning. The resulting knowledge
structure leads to a natural assessment methodology in which the expert
first specifies a qualitative argument from evidence to conclusion. Next
the expert specifies a series of premises on which the argument is
based. Invalidating any of these premises would disrupt the causal link
between evidence and conclusion. The final step is the assessment of the
strength of the argument, in the form of degrees of belief for the
premises underlying the argument. The expert may also explicitly adopt
assumptions affecting the strength of evidential arguments. A
higher-level `metareasoning' process is described, in which assumptions
underlying the strength and direction of evidential arguments may be
revised in response to conflict
The reactor safety study
- rasmussen