41.
Two Topologies Are Better Than One
Partially ordered sets and metric spaces are used in studying semantics in Computer
Science. Sets with both these structures are hence of particular interest. The
partial metric spaces introduced by Matthews are an attempt to bring these ideas
together in a single axiomatic framework.
We consider an appropriate context in which to consider these spaces is as a bitopological
space, i.e. a space with two (related) topologies. From this starting point, we
cover the groundwork for a theory of partial metric spaces by generalising ideas from
topology and metric spaces.
For intuition we repeatedly refer to the real line with the usual ordering and metric
as a natural example....
42.
Inductively Generated Formal Topologies
- Thierry Coquand,Giovanni Sambin,Jan Smith,Silvio Valentini
Formal topology aims at developing general topology in intuitionistic
and predicative mathematics. Many classical results of general topology
have been already brought into the realm of constructive mathematics by
using formal topology and also new light on basic topological notions was
gained with this approach which allows distinction which are not sensible
in classical topology.
Here we give a systematic exposition of one of the main tools in formal
topology: inductive generation. In fact, many formal topologies can be
presented in a predicative way by an inductive generation and thus their
properties can be proved inductively. We show however that some natural
complete Heyting algebra cannot be inductively defined.
Contents
1 The...
43.
Ambiente para otimização de redes multimídia utilizando algoritmo genético
- Ricardo Martins Lemos
O objetivo deste trabalho é desenvolver ferramentas computacionais para a obtenção de um ambiente que permita a busca de topologias ótimas para interligação dos nós de uma rede multimídia. Isto será realizado com base na quantidade de tráfego a ser transportado, nas informações de custo e confiabilidade dos tipos de interfaces de transmissão disponíveis e na distância entre os nós. Topologia ótima é aquela que pelo menor custo atenda aos requisitos de confiabilidade mínima, ocupação máxima de enlaces, atraso de enfileiramento máximo e perda máxima de pacotes definidos pelo planejador. Os cálculos para atraso e perda foram baseados nos modelos...
44.
Trivialidad definible de familias de aplicaciones definibles en estructuras o-minimales
- Escribano Martínez, Jesús
El objetivo de la memoria es estudiar la trivialidad de sumersiones (y pares de sumersiones) dentro de la categoría o-minimal. Este es un problema clásico de la Topología Diferencial y con numerosas aplicaciones en la Teoría de singularidades. Para este objetivo ampliamos a la categoría o-minimal diversas construcciones de la geometría semi-algebraica, como el espectro real. Se construye entonces el espectro definible y se relaciona con las familias de objetos definibles. A continuación se estudia un teorema de aproximación de funciones diferenciables definibles por funciones con una clase de diferenciabilidad más alta. Utilizando este resultado de aproximación, y los resultados...
45.
Simulating complex SCI topologies
- Geir Horn,John W. Bothner,Svein Linge,Ernst H. Kristiansen,Ystein Gran Larsen
This paper shows how rather complex SCI
topologies might be constructed and simulated using our
present SCI simulator. We also present the serial HIC
technology developed in the European OMI/HIC project. A
HIC network may be used to transport SCI packets, and our
new HIC network simulator under development is presented
as it is intended to simulate such HIC networks.
Keywords--- Scalable Coherent Interface (SCI), Simulations,
HIC Technology, Serial links, Topologies
I. Introduction
N
ODES in an SCI topology are designed to form
ringlets. However, ringlet structures are sensitive to
hardware failures, and their peak load is limited; they are
not truly scalable [1]. Having access to switches that enable
traffic to be directed from...
46.
Flexible Distributed Process Topologies for Enterprise Applications
- Christoph Hartwich
Enterprise applications can be viewed as topologies of distributed processes that access
business data objects stored in one or more transactional datastores. There are several wellknown
topology patterns that help to integrate different subsystems or to improve
nonfunctional properties like scalability, fault tolerance, or response time. Combinations of
multiple patterns lead to custom topologies with the shape of a directed acyclic graph
(DAG). These topologies are hard to build on top of existing middleware and even harder
to adapt to changing requirements. In this paper we present the principles of an enterprise
application architecture that supports a wide range of custom topologies. The architecture
decouples application code, process...
47.
O modelo esférico e a hipótese topológica sobre transições de fase
- Teixeira, Ana Carolina Ribeiro
O presente trabalho apresenta a verificação da Hipótese Topológica (HT) sobre o Modelo Esférico ferromagnético. Esta hipótese sugere uma nova abordagem a transições de fase termodinâmicas (TF), apoiada na idéia recentemente proposta de que o mecanismo na origem das transições de fase seja uma quebra de difeomorficidade adequada do espaço de configurações dos sistemas no ponto da transição. Embora se saiba que nem toda transição topológica (TT) esteja ligada a uma TF, foi provado recentemente para uma classe de modelos que TTs são necessárias na presença de transições de fase de primeira e segunda ordem, pelo menos. Da termodinâmica do...
48.
Evolving Neural Networks through Augmenting Topologies
- Kenneth O. Stanley,Risto Miikkulainen
An important question in neuroevolution is how to gain an advantage from evolving neural network topologies along with weights. We present a method, NeuroEvolution of Augmenting Topologies (NEAT) that outperforms the best fixed-topology method on a challenging benchmark reinforcement learning task. We claim that the increased efficiency is due to (1) employing a principled method of crossover of different topologies, (2) protecting structural innovation using speciation, and (3) incrementally growing from minimal structure. We test this claim through a series of ablation studies that demonstrate that each component is necessary to the system as a whole and to each other....
49.
Efficient Evolution of Neural Network Topologies
- Kenneth O. Stanley,Risto Miikkulainen
Neuroevolution, i.e. evolving artificial neural networks with genetic algorithms, has been highly effective in reinforcement learning tasks, particularly those with hidden state information. An important question in neuroevolution is how to gain an advantage from evolving neural network topologies along with weights. We present a method, NeuroEvolution of Augmenting Topologies (NEAT) that outperforms the best fixed-topology methods on a challenging benchmark reinforcement learning task. We claim that the increased efficiency is due to (1) employing a principled method of crossover of different topologies, (2) protecting structural innovation using speciation, and (3) incrementally growing from minimal structure. We test this claim...
50.
Evolving Neural Networks through Augmenting Topologies
- Kenneth O. Stanley
An important question in neuroevolution is how to gain an advantage from evolving neural network topologies along with weights. We present a method, NeuroEvolution of Augmenting Topologies (NEAT), which outperforms the best fixed-topology method on a challenging benchmark reinforcement learning task. We claim that the increased efficiency is due to (1) employing a principled method of crossover of different topologies, (2) protecting structural innovation using speciation, and (3) incrementally growing from minimal structure. We test this claim through a series of ablation studies that demonstrate that each component is necessary to the system as a whole and to each other....
51.
Efficient Evolution of Neural Network Topologies
- Kenneth O. Stanley,Risto Miikkulainen
Neuroevolution, i.e. evolving artificial neural networks
with genetic algorithms, has been highly effective in reinforcement
learning tasks, particularly those with hidden
state information. An important question in neuroevolution
is how to gain an advantage from evolving neural network
topologies along with weights. We present a method,
NeuroEvolution of Augmenting Topologies (NEAT) that
outperforms the best fixed-topology methods on a challenging
benchmark reinforcement learning task. We claim that
the increased efficiency is due to (1) employing a principled
method of crossover of different topologies, (2) protecting
structural innovation using speciation, and (3) incrementally
growing from minimal structure. We test this claim
through a series of ablation studies that demonstrate that
each component is necessary to...
52.
Analysing Topologies of Transitive Trust
- Audun Jsang,Elizabeth Gray,Michael Kinateder
Transacting and interacting through computer networks makes it difficult to use traditional methods for establishing trust between parties. Creating substitutes by which people, organisations and software agents can derive trust in others through computer networks requires computerised analysis of trust topologies. This paper describes diverse dimensions of trust that are needed for analysing trust topologies, and provides a notation with which to express trust relationships in terms of these dimensions. The result is a simple way of specifying topologies of trust from which derived trust relationships can be automatically and securely computed.
53.
Deriving Non-Hierarchical Process Topologies
- Ricardo Pea,O Rubio,Clara Segura,Informticos Programacin
Eden is a parallel functional language which extends Haskell with new expressions to define and instantiate
processes. These extensions allow the easy definition of parallel process topologies as higher order functions.
54.
Arbitrary Topologies
- Xianfeng Gu
Surface parametrization is a fundamental problem in computer graphics. It
is essential for operations such as texture mapping, texture synthesis, interactive
3D painting, remeshing, multi-resolution analysis and mesh compression.
Conformal parameterization, which preserves angles, has many nice properties
such as having no local distortion on textures, and being independent of triangulation
or resolution. Existing conformal parameterization methods partition
a mesh into several charts, each of which is then parametrized and packed to
an atlas. These methods su#er from limitations such as di#culty in segmenting
the mesh and artifacts caused by discontinuities between charts.
55.
Designing Distance-Preserving Fault-Tolerant Topologies
- Sitarama Swamy,Kocherlakota Disseertation
DESIGNING DISTANCE-PRESERVING FAULT-TOLERANT TOPOLOGIES
By
Sitarama Swamy Kocherlakota
The objective of designing a fault-tolerant communication network (CN) is to maintain its
functionality in the presence of failures. The approaches taken in designing fault-tolerant systems
can be classified, in general, by the functionality criteria. There are two functionality criteria that
have received much attention. First, preserving a topology, which considers a CN functional as long
as a desired topology is contained in the system. Second, preserving a property, which considers
a CN functional as long as a given topological property is satisfied by the underlying topology. In
this dissertation, we study the second functionality criterion. Specifically, designing fault-tolerant
topologies that preserve...
56.
Interconnection Topologies for Parallel Processing Systems
- Gabriele Kotsis
Introduction
The growing demand for more computing power at increasing speed in many scientific
and engineering applications made it necessary to develop advanced computer architectures
based on the concept of parallel processing. In general a parallel computer
system consists of various processing and memory units and other (shared) resources.
A critical issue in design and analysis of parallel systems is the way in which the system
components are connected together, since this interconnection network determines the
performance of the whole system [Bhuy 87]. The network topology, defined as the abstract
representation of the connections in the network [?], is a key factor in determing
a suitable architectural structure. A...
57.
Generating Realistic ISP-Level Network Topologies
- Michael Piringer,Jens Schmitt,Ralf Steinmetz,Oliver Heckmann
Simulations are an important tool in network
research. As the selected topology often influences the outcome of
the simulation, realistic topologies are needed to produce realistic
simulation results. Using several similarity metrics to compare
artificially generated topologies with real world topologies this letter
gives hints how to use the wide-spread topology generators
BRITE, TIERS and GT-ITM to create realistic topologies.
58.
Random Finite Topologies and Their Thresholds
- C. F. Maclean
For each integer n, there is a natural family of probability distributions
on the set of topologies on a set of n elements, parameterised
by an integer variable, m. We will describe how these are constructed
and analysed, and find threshold functions (for m in terms of n) for
various topological properties.
1
An earlier version of this article appeared as BRIMS Technical Report HPL-BRIMS9718
2
Cambridge University. Based on work completed as a summer student at BRIMS,
1997.
3
BRIMS, HP Labs, Stoke Gifford, Bristol BS12 6QZ.
Suppose we have a set, S, and we want to pick a topology on S at random.
What do we mean when we say "at...
59.
Efficient Lookup on Unstructured Topologies
- Ruggero Morselli,Bobby Bhattacharjee,Michael A. Marsh,Aravind Srinivasan
We present LMS, a protocol for efficient lookup on unstructured
networks. Our protocol uses a virtual namespace without imposing
specific topologies. It is vastly more efficient than existing
lookup protocols for unstructured networks, and thus is an attractive
alternative for applications in which the topology cannot
be structured as a DHT.
60.
Metrics and Topologies for Geographic Space
- Michael F Worboys
This paper is motivated by the requirement for computer-based representations of geospatial information that parallel the structure of geographic space that humans apprehend. It is well known that such representations do not satisfy the conditions of a metric space. We consider distance and proximity relationships between entities in geographic spaces, and discuss some of the ways in which useful representations take us beyond mathematical metric spaces. We show that even though the underlying space does not satisfy the conditions for a metric space, it is still possible to induce neighbourhood topologies upon it.
