Showing papers in "Future Generation Computer Systems in 1997"

TL;DR: A new interest-measure for rules which uses the information in the taxonomy is presented, and given a user-specified “minimum-interest-level”, this measure prunes a large number of redundant rules.

TL;DR: This paper describes the use of decision tree and rule induction in data-mining applications and presents a synopsis of some major state-of-the-art tree andrule mining methodologies, as well as some recent advances.

TL;DR: In this paper, the authors discuss two classes of approaches for data mining with neural networks: rule extraction and directly learning simple, easy-to-understand networks, and argue that, given the current state-of-the-art, neural-network methods deserve a place in the tool boxes of data-mining specialists.

TL;DR: An overview of this growing multi-disciplinary research area of data mining and knowledge discovery in databases is provided, outline the basic techniques, and provide brief coverage of how they are used in some applications.

TL;DR: This paper proposes a methodology for comparing clustering methods based on the quality of the result and the performance of the execution, and applied it to several known clustering method: FastClust, Autoclass, Relational data analysis, and Kohonen nets.

TL;DR: Comparing three approaches to machine learning that have developed largely independently: classical statistics, Vapnik's statistical learning theory, and computational learning theory concludes that statisticians and data miners can profit by studying each other's methods and using a judiciously chosen combination of them.

TL;DR: This work examines various important concepts and approaches that are used for modelling a target attribute by other attributes in the data and contrasts their strengths.

TL;DR: A cryptographic coprocessor is described for certifying outcomes of software programs and allowing a third party who trusts the secure components of the system to verify that a specied program actually executed and produced a claimed output.

TL;DR: The main features of particles paradigm proposed: simplicity, decomposition ability, and built-in message-passing way of communication, make it attractive as a universal approach for mapping different problems on parallel platform.

TL;DR: It is shown that for a particular class of dummy insertion strategies, a Viterbi decoder can fairly reliably distinguish dummy fetches from real instruction fetches and that in such a case, it is fundamentally not possible to enhance the security of the implementation by keeping the strategy for generating dummy Fetches secret to the attacker.

TL;DR: The usage of object-oriented technologies are presented to implement a generic smart card operating system and to provide a card object adapter to access smart card services from distributed object- oriented information systems based on CORBA architecture.

TL;DR: It is found that the proposed methodology can be especially useful for load balancing in asymmetric processor topologies, and therefore is of importance for workload balancing in workstation clusters.

TL;DR: CAMEL is an interactive environment designed to support the development of high performance applications in science and engineering that offers the computing power of a highly parallel computer, hiding the architecture issues from a user.

TL;DR: It is indicated why it is important that tools to handle different aspects of the K DD effort need to be integrated and several enhancements to the underlying database systems that may significantly benefit the KDD process are suggested.

TL;DR: A new approach for load balancing in high-performance parallel database systems is outlined, based on the architectural model of a distributed system where the communication between database servers is implemented by means of the technology of Interface Definition Languages.

TL;DR: An overview of the Vipar project is presented and the libraries developed to support the first phase of the work which is a tool to describe parallel visualization modules are described.

TL;DR: This paper summarizes some of the most important observations about HPC as of late 1996, in particular the continued dominance of the world market in HPC by the US, the market penetration by commodity microprocessor based systems, and the growing industrial use of super-computers.

TL;DR: A modified coherency protocol with static cache line ownership is proposed that allows locality to be exploited to at least the level of a conventional protocol, but without the unnecessary serialisation and network transactions this usually causes.

TL;DR: Performance improvement that can be obtained by increasing the size of a distributed memory MIMD system depends on the characteristics of the workload as well as the parameters that characterize the communication speed of the parallel system.

TL;DR: Initial performance comparisons of two matrix algorithms using MPI-DDL, HPF, and direct MPI implementations are presented and a new point of view is presented that regards distributed data as a programming unit.

TL;DR: This paper introduces location-independent transformations (LITs) as a general strategy for implementing distributed feed forward networks that use dynamic topologies (dynamic ANNs) efficiently in parallel hardware.

TL;DR: This paper shows that a task compatibility graph can be generated by analyzing certain compatible relations between task module pairs of a given task flow graph, and shows that for three special types of task compatibility graphs, optimal solution can be obtained in polynomial time.

TL;DR: This paper explores some of the issues for a intelligent storage device required to provide a uniform interface regardless of implementation and which also directly supports the basic need of scalability.

TL;DR: A general method to isolate the loop-body of a nested loop for all iterations in a particular polyhedral set by successively partitioning execution sets of individual DO-loops in the loop, thereby partitioning the iteration space according to the linear inequalities that define thepolyhedral set.

TL;DR: A simulation-based study shows that the Generalized Relay Mechanism is a very competitive alternative in comparision with simply incorporating more hard disks and enjoys an advantage in terms of cost.

TL;DR: The key challenges for very large workstation networks are identified, and implementation techniques at system software level to overcome these problems are described.