Bayesian Networks is a popular tool for representing uncertainty knowledge in artificial intelligence fields. Learning BNs from data is helpful to understand the casual relation between variables. But Learning BNs is ...
Bayesian Networks is a popular tool for representing uncertainty knowledge in artificial intelligence fields. Learning BNs from data is helpful to understand the casual relation between variables. But Learning BNs is a NP hard problem. This paper presents an immune genetic algorithm for learning Markov equivalence classes, which combining dependency analysis and search-scoring approach together. Experiments show that the immune operators can constrain the search space and improve the {1.al performance.
Hierarchy is a remedy way to reduce the demanding complexity of model-based diagnosis. In this paper, an approach to diagnosis of discrete-event systems in a hierarchical way is proposed, inspired by the concept "...
详细信息
Hierarchy is a remedy way to reduce the demanding complexity of model-based diagnosis. In this paper, an approach to diagnosis of discrete-event systems in a hierarchical way is proposed, inspired by the concept "D-holon" and the concept "Silent Closure" presented in the literatures recently. Each extended silent closure can be seen as a special type of D-holons, called SCL-D-holon. Every hierarchical level is an SCL-D-holon built off line. When on line diagnosing a discrete-event system, only related SCL-D-holons will be called instead of all the SCL-D-holons generally, thus the space complexity is reduced. In comparison to on line creating silent closures, the efficiency is improved as well.
Proper ontology definition is the prerequisite for efficient knowledge acquisition. For the complex knowledge that could not be described by simple binary relation, we advocated a methodology for aggregated knowledge ...
详细信息
ISBN:
(纸本)9781424430536;9780769531519
Proper ontology definition is the prerequisite for efficient knowledge acquisition. For the complex knowledge that could not be described by simple binary relation, we advocated a methodology for aggregated knowledge acquisition, describing how to define the ontology for such aggregated knowledge concept and how to acquire knowledge basing on such definition. Experiment shows that this methodology is effective in automatic knowledge acquisition from Chinese free text.
We present a novel object localization approach based on the global structure constraint model (GSC) and optimal algorithm. In GSC, Objects are described as constellations of points satisfied with their specific globa...
详细信息
ISBN:
(纸本)9781424431977;9780769531618
We present a novel object localization approach based on the global structure constraint model (GSC) and optimal algorithm. In GSC, Objects are described as constellations of points satisfied with their specific global structure constraints. The spatial relations among all the patches having stable color information and their representative color information around patches are encoded. Then, the searching algorithm, i.e. genetic algorithm, is used to locate target objects in images by finding out the exactly matched position. In the experiment, we tested the approach on a collection of human face images and the results demonstrated the approach is simple, effective and efficient.
Attribute discretization is one of the key issues for the Rough Set theory. First, a method is proposed to compute an initial cut points set. The indistinguishable relation of decision tables did not change, and the n...
详细信息
Attribute discretization is one of the key issues for the Rough Set theory. First, a method is proposed to compute an initial cut points set. The indistinguishable relation of decision tables did not change, and the number of elements in the initial cut points set was reduced. Then, the cut point information entropy was defined to measure the importance of a cut point. Finally, an attribute discretization algorithm based on the Rough Set and information entropy was proposed. The consistence of decision tables did not change, and the mixed decision table was considered, which contains continuous and discrete attributes. The experimental results show that this algorithm is effective and is competent for processing the large-scale datasets.
DNA-binding proteins play an important role invarious intra-and extra-cellular *** key in theprotein is DNA-binding region also called DNA-bindingdomain(DBD).However,it is hard to search the DBDsby means of homology s...
详细信息
DNA-binding proteins play an important role invarious intra-and extra-cellular *** key in theprotein is DNA-binding region also called DNA-bindingdomain(DBD).However,it is hard to search the DBDsby means of homology search or hidden Markov modelsbecause of a wide variety of the *** this work,we develop a kernel-based machine learning method bycombination of multiple "l-vs-l" binary classifiers forDNA binding domain *** result shows that93.73% accuracy is achieved for multicategory classifierand no less than 90% accuracy for each binary *** comparison,our classifier performs better than othermachine learning methods.
Aiming at the dynamic policy trigger problem in policy based network management, this thesis first introduces the uncertain information generated by one-point coverage random sets theory and also the unified framework...
详细信息
Aiming at the dynamic policy trigger problem in policy based network management, this thesis first introduces the uncertain information generated by one-point coverage random sets theory and also the unified framework of fuzzy information. On the basis of which, according to the feature of policy based network management, a model used to update the predicted information using present known information, experience, and knowledge is designed. At the same time, this thesis put forward that, take the semblance of policy as constant probability in product space conditional event algebra, so as to execute information fusion of multiple sets of related policy. In the end, compared with the policy based inference system and then analyze the result.
A framework of hiberarchy trusted network basedon the grade division was put forward,and thepartition rules of trusted attributes as well as themethods of grade division were explained in ***,the potential application...
详细信息
A framework of hiberarchy trusted network basedon the grade division was put forward,and thepartition rules of trusted attributes as well as themethods of grade division were explained in ***,the potential applications of this frameworkin trusted network were discussed,and the accessprocedure of terminals in hiberarchy trusted networkwas given based on the existing research *** show that by the proposed framework,notonly the safety and reliability of network can beensured,but also the flexibility of ways to accessnetwork is strengthened,and that the framework canprovide supports for interoperability of differentequipment suppliers.
Internet is a complex network with the characteristic of self-organized criticality. The cascading dynamics of Internet are presented and two reasons are pointed out, which may cause cascading failures. Different from...
详细信息
Internet is a complex network with the characteristic of self-organized criticality. The cascading dynamics of Internet are presented and two reasons are pointed out, which may cause cascading failures. Different from betweenness centrality, a congestion function to represent the extent of congestion is proposed. By introducing the concept of "delay time", the correlation between permanent removing and non-removing is built. And a new evaluation function of network efficiency based on congestion function is given in order to measure the damage of cascading failures. Moreover some effects of network structure and size, delay time on congestion propagation are also investigated, and cascading process composed of three phases and some factors affecting cascade propagation are uncovered.
暂无评论