this paper presents a method for enhancing the performance of current clustering algorithms; the method is based on Particle Swarm Optimization techniques. Namely, a preprocessing step aims at bringing rdquocloserrdqu...
详细信息
this paper presents a method for enhancing the performance of current clustering algorithms; the method is based on Particle Swarm Optimization techniques. Namely, a preprocessing step aims at bringing rdquocloserrdquo objects which are likely to belong to the same cluster, while increasing the distance between objects likely to belong to different clusters. Experimental results show significantly improved performance for further clustering procedures especially when non-spherical clusters are involved.
Recent years have seen tremendous progress in the capabilities of software model checkers. these tools are the result of combined efforts by different communities, most notably model checking, abstract interpretation,...
详细信息
Recent years have seen tremendous progress in the capabilities of software model checkers. these tools are the result of combined efforts by different communities, most notably model checking, abstract interpretation, program semantics, and decision procedures. In this tutorial, we will introduce the fundamentals of software model checking, and show how the different origins of the field combine to a common agenda.
BBUFs (Babes Bolyai university file system) is a peer to peer distributed file system with a new lookup mechanism. the file system is designed for ordinary Unix machines that are IPv6 capable. In this paper we propose...
详细信息
BBUFs (Babes Bolyai university file system) is a peer to peer distributed file system with a new lookup mechanism. the file system is designed for ordinary Unix machines that are IPv6 capable. In this paper we propose a new lookup mechanism based on IPv6, and its newly defined addressing scheme called "anycast address". the advantages and disadvantages of this approach are also discussed.
the mathematical backbone of this article is formed by three classical formulas of Wallis: his product formula for π, an inequality implying the product formula in the limit, and a related definite integral involving...
详细信息
the mathematical backbone of this article is formed by three classical formulas of Wallis: his product formula for π, an inequality implying the product formula in the limit, and a related definite integral involving powers of the sine function. For the latter we present various evaluations to illustrate recent algorithmic developments. In the main part of the article we automatically refine the bounds in Wallis' inequality by using methods based on Collins' cylindrical algebraic decomposition.
We present an algorithm for solving Diophantine equations which are linear in the variables, but non-linear in one parameter. We are looking for the pointwise solutions, i.e., the solutions for the unknowns in depende...
详细信息
We present an algorithm for solving Diophantine equations which are linear in the variables, but non-linear in one parameter. We are looking for the pointwise solutions, i.e., the solutions for the unknowns in dependence of the value of the parameter. Solving Diophantine equations is central to computingthe data dependences of certain codes (loops with certain array accesses) which often occur in scientificcomputing. Our algorithm enables the computation of data dependences in more general situations than is possible with current algorithms.
the family of default logics formalize the default reasoning using nonmonotonic inference rules called defaults. In this paper we propose a uniform abstract characterization of credulous default inference associated t...
详细信息
the family of default logics formalize the default reasoning using nonmonotonic inference rules called defaults. In this paper we propose a uniform abstract characterization of credulous default inference associated to all versions (classical, justified, constrained, rational) of propositional default logic using the credulous default sequent calculi. these axiomatic systems combine sequent and anti-sequent calculus rules for propositional logic with reduction rules specific to the application of the defaults.
We propose a new technique for identifying frequent patterns that occur in large transactional data sets using Boolean algebras. Our approach, which involves extending the notion of support from sets of attributes to ...
详细信息
ISBN:
(纸本)9781538626276
We propose a new technique for identifying frequent patterns that occur in large transactional data sets using Boolean algebras. Our approach, which involves extending the notion of support from sets of attributes to Boolean functions is more expressive than the standard approach that amounts to a search for conjunctive patterns. An efficient algorithm based on computing supports for minterms is presented. Also, we discuss an application of this algorithm to finding independent sets in graphs.
In this paper we express Babes Bolyai University File System lookup mechanism by using π-calculus. We describe the lookup process in a peer-to-peer decentralized system, how a request message is forwarded from a clie...
详细信息
In this paper we express Babes Bolyai University File System lookup mechanism by using π-calculus. We describe the lookup process in a peer-to-peer decentralized system, how a request message is forwarded from a client to a system node, and how the response is replied. the formally specified protocol is verified by using the Mobility Workbench model-checker.
Many complex systems need in an explicit or in an implicit form the ontologies. One of the major problems consist in that the need to obtain an ontology from some others using appropriate operations. In many domains t...
详细信息
Many complex systems need in an explicit or in an implicit form the ontologies. One of the major problems consist in that the need to obtain an ontology from some others using appropriate operations. In many domains the ontologies are in some way related. the ontology derivation allows obtaining one ontology from another. Paper analyzes some related previous works and suggests some primary formalism that allows deriving one ontology from other one.
Service oriented architectures become more and more popular withthe emergence and consolidation of new paradigms such as Clouds and Grids. In this context scheduling becomes an important and difficult problem as serv...
详细信息
Service oriented architectures become more and more popular withthe emergence and consolidation of new paradigms such as Clouds and Grids. In this context scheduling becomes an important and difficult problem as services hide their actual implementation, requirements or efficiency and are spread across multiple institutions and geographical locations. this paper proposes a distributed scheduling approach based on agents where each agent handles a certain domain and manages independently the policies inside it. In its frame several scheduling algorithms are also studied, compared and some results are given.
暂无评论