Updated Graph Research Articles

Graph optimization for dimensionality reduction with sparsity constraints

Graph-based dimensionality reduction (DR) methods play an increasingly important role in many machine learning and pattern recognition applications. In this paper, we propose a novel graph-based learning scheme to conduct Graph Optimization for Dimensionality Reduction with Sparsity Constraints (GODRSC). Different from most of graph-based DR methods where graphs are generally constructed in advance, GODRSC aims to simultaneously seek a graph and a projection matrix preserving such a graph in one unified framework, resulting in an automatically updated graph. Moreover, by applying an l1 regularizer, a sparse graph is achieved, which models the ''locality'' structure of data and contains natural discriminating information. Finally, extensive experiments on several publicly available UCI and face databases verify the feasibility and effectiveness of the proposed method.

A CACHING TECHNIQUE FOR OPTIMIZING AUTOMATED SERVICE DISCOVERY

The development of sophisticated technologies for service-oriented architectures (SOA) is a grand challenge. A promising approach is the employment of semantic technologies to better support the service usage cycle. Most existing solutions show significant deficits in the computational performance, which hampers the applicability in large-scale SOA systems. We present an optimization technique for automated service discovery — one of the central operations in semantically enabled SOA environments — that can ensure a sophisticated performance while maintaining a high retrieval accuracy. The approach is based on goals that formally describe client objectives, and it employs a caching mechanism for enhancing the computational performance of a two-phased discovery framework. At design time, the suitable services for generic and reusable goal descriptions are determined by semantic matchmaking. The result is captured in a continuously updated graph structure that organizes goals and services with respect to the requested and provided functionalities. This is exploited at runtime in order to detect the suitable services for concrete client requests with minimal effort. We formalize the approach within a first-order logic framework, and define the graph structure along with the associated storage and retrieval algorithms. An empirical evaluation shows that significant performance improvements can be achieved.

Cooperative Search by UAV Teams: A Model Predictive Approach using Dynamic Graphs

A receding-horizon cooperative search algorithm is presented that jointly optimizes routes and sensor orientations for a team of autonomous agents searching for a mobile target in a closed and bounded region. By sampling this region at locations with high target probability at each time step, we reduce the continuous search problem to a sequence of optimizations on a finite, dynamically updated graph whose vertices represent waypoints for the searchers and whose edges indicate potential connections between the waypoints. Paths are computed on this graph using a receding-horizon approach, in which the horizon is a fixed number of graph vertices. To facilitate a fair comparison between paths of varying length on nonuniform graphs, the optimization criterion measures the probability of finding the target per unit travel time. Using this algorithm, we show that the team discovers the target in finite time with probability one. Simulations verify that this algorithm makes effective use of agents and outperforms previously proposed search algorithms. We have successfully hardware tested this algorithm in two small unmanned aerial vehicles (UAVs) with gimbaled video cameras.

Publishing anonymous survey rating data

We study the challenges of protecting privacy of individuals in the large public survey rating data in this paper. Recent study shows that personal information in supposedly anonymous movie rating records are de-identified. The survey rating data usually contains both ratings of sensitive and non-sensitive issues. The ratings of sensitive issues involve personal privacy. Even though the survey participants do not reveal any of their ratings, their survey records are potentially identifiable by using information from other public sources. None of the existing anonymisation principles (e.g., k-anonymity, l-diversity, etc.) can effectively prevent such breaches in large survey rating data sets. We tackle the problem by defining a principle called $${(k,\epsilon)}$$ -anonymity model to protect privacy. Intuitively, the principle requires that, for each transaction t in the given survey rating data T, at least (k ? 1) other transactions in T must have ratings similar to t, where the similarity is controlled by $${\epsilon}$$ . The $${(k,\epsilon)}$$ -anonymity model is formulated by its graphical representation and a specific graph-anonymisation problem is studied by adopting graph modification with graph theory. Various cases are analyzed and methods are developed to make the updated graph meet $${(k,\epsilon)}$$ requirements. The methods are applied to two real-life data sets to demonstrate their efficiency and practical utility.

Network Analysis of World Subway Systems Using Updated Graph Theory

This paper demonstrates that network topologies play a key role in attracting people to use public transit; ridership is not solely determined by cultural characteristics (North American versus European versus Asian) or city design (transit oriented versus automobile oriented). The analysis considers 19 subway systems worldwide: those in Toronto, Ontario, Canada; Montreal, Quebec, Canada; Chicago, Illinois; New York City; Washington, D.C.; San Francisco, California; Mexico City, Mexico; London; Paris; Lyon, France; Madrid, Spain; Berlin; Athens, Greece; Stockholm, Sweden; Moscow; Tokyo; Osaka, Japan; Seoul, South Korea; and Singapore. The relationship between ridership and network design was studied by using updated graph theory concepts. Ridership was computed as the annual number of boardings per capita. Network design was measured according to three major indicators. The first is a measure of transit coverage and is based on the total number of stations and land area. The second relates to the maximum number of transfers necessary to go from one station to another and is called directness. The third attempts to get an overall view of transfer possibilities to travel in the network to appreciate a sense of mobility; it is termed connectivity. Multiple-regression analysis showed a strong relationship between these three indicators and ridership, achieving a goodness of fit (adjusted R2 value) of .725. The importance of network design is significant and should be considered in future public transportation projects.

Linking the results of the regional key comparison APMP.AUV.V-K1 to those of the CIPM key comparison CCAUV.V-K1

During 1996 and 1997, eight national metrology institutes (NMI) took part in a vibration accelerometer comparison, identifier APMP.AUV.V-K1. Two NMIs ultimately withdrew from the comparison and the results of the remaining six NMIs have been approved by the CCAUV. Four NMIs subsequently took part in the 2001 CIPM key comparison for the same quantity, identifier CCAUV.V-K1. The results of these four CIPM participants have been used to link the results of the remaining two NMIs to the results in the CIPM key comparison using the reference frequency of 160 Hz. The CCAUV nominated the PTB to propose the methodology for the link and subsequently approved the linked results as presented in this report. The degrees of equivalence between each result and the key comparison reference value (KCRV), and between each NMI have been calculated and the results are given in the form of a matrix and graph for six NMIs. As two results from the APMP can now be linked to the published CCAUV.V-K1 comparison, the updated graph for the key comparison database is also given.Main text.To reach the main text of this paper, click on Final Report.The final report has been peer-reviewed and approved for publication by the CCAUV, according to the provisions of the Mutual Recognition Arrangement (MRA).

중복 데이터베이스 시스템에서 갱신그래프를 이용한 동시성제어

중복 데이터베이스 시스템에서 갱신그래프를 이용한 동시성제어

Elements of a theory of computer simulation I: Sequential CA over random graphs

This paper is a first step in the development of mathematical foundations for a theory of simulation. We employ sequentially updated cellular automata (sCA) over arbitrary graphs as a paradigmatic framework. In the development of the theory, we focus on the properties of causal dependencies among local mappings in a simulation. Let Y be a graph in which (a) each vertex i has a state x i 0, 1 and (b) there exists a local map ƒ i defined on the states of the Y-neighbors and x i that returns the state of y i 0, 1. Suppose i 1, …, i n is a permutation of the Y-vertices, then the order of application of the local maps ƒ ik induces a sequential (or asynchronous) cellular automaton (sCA) over Y. In this paper we introduce a mapping between the base graph Y, over which the sCA is defined and which represents the mutual dependencies of the local maps, and the update graph U(Y), whose vertices are permutations of all Y-vertices. Two permutations π 1, π 2 are adjacent in U( Y) if they differ in exactly two consecutive coordinates, { i k, i k+1|, and { i k, i k+1 ∉ eY. U( Y) represents a conceptual framework which allows to determine equivalence classes of sCA for a given graph X, independent of the family of local maps ( ƒ i) 1 ⩽ i ⩽ n . We consider: (a) U as a random variable over G n,p and analyze the induced random graph U( G n,p ), the update graph, and show (b) that U exhibits properties of a covariant functor.

An interactive multimedia business game

The playing of computer-based business games has changed little with recent technological improvements. Teams make decisions which are written down on paper and handed to a tutor who then has to enter these into the computer. To make it a richer and more real-life experience we have brought a business simulation into a multimedia environment. In order to create a sense of the passing of time the simulation is continuous, and players can enter input values throughout the game. Important parameters, such as product sales, are displayed as continuously updated graphs. Background information such as news items is provided in multimedia form. Players also have direct access to online tools, such as spreadsheets, from the game environment. We describe the prototype built to illustrate our approach.

SLUG: A connectionist architecture for inferring the structure of finite-state environments

Consider a robot wandering around an unfamiliar environment, performing actions and sensing the resulting environmental states. The robot's task is to construct an internal model of its environment, a model that will allow it to predict the consequences of its actions and to determine what sequences of actions to take to reach particular goal states. Rivest and Schapire (1987a, 1987bs Schapire, 1988) have studied this problem and have designed a symbolic algorithm to strategically explore and infer the structure of “finite state” environments. The heart of this algorithm is a clever representation of the environment called an update graph. We have developed a connectionist implementation of the update graph using a highly-specialized network architecture. With back propagation learning and a trivial exploration strategy—choosing random actions—the network can outperform the Rivest and Schapire algorithm on simple problems. Perhaps the most interesting consequence of the connectionist approach is that, by relaxing the constraints imposed by a symbolic description, it suggests a more general representation of the update graph, thus allowing for greater flexibility in expressing potential solutions.

Discovering the Structure of a Reactive Environment by Exploration

Consider a robot wandering around an unfamiliar environment, performing actions and observing the consequences. The robot's task is to construct an internal model of its environment, a model that will allow it to predict the effects of its actions and to determine what sequences of actions to take to reach particular goal states. Rivest and Schapire (1987a,b; Schapire 1988) have studied this problem and have designed a symbolic algorithm to strategically explore and infer the structure of “finite state” environments. The heart of this algorithm is a clever representation of the environment called an update graph. We have developed a connectionist implementation of the update graph using a highly specialized network architecture. With backpropagation learning and a trivial exploration strategy — choosing random actions — the connectionist network can outperform the Rivest and Schapire algorithm on simple problems. Our approach has additional virtues, including the fact that the network can accommodate stochastic environments and that it suggests generalizations of the update graph representation that do not arise from a traditional, symbolic perspective.

On the use of colours for presenting power system connectivity information

A coloring concept is presented which shows the voltage level and the status of the power system element and allows the identification of the existing connectivity units. Connectivity units may be nodes, network groups (network islands), network districts (sets of interconnected nodes on the same level), and power source groups (elements fed by a common source point). The task of connectivity unit coloring is converted to the task of coloring a graph. This enables graph theory to be applied. An update algorithm is provided to update graph colors after vertex-edge changes caused by switch indication changes. Issues in implementing such a power system coloring function within a full graphic based energy management system are discussed. A typical example is the trade-off between software effort, computational complexity (performance), and the number of colors required.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Graph Property Update Algorithms and Their Appligation to Distance Matrices

Algorithms are considered to update graph properties after small changes have been made to a graph. Such update algorithms can use the approach either of directly determining the properties from the new graph or of actually updating the properties of the previous graph. Unfortunately, the execution times for the update algorithms developed by these two approaches often have the same worst-case order. One approach to comparing algorithms with equal worst-case orders is to count the frequency of execuuon for key operations, called active operations. For a class of algorithms that includes most update algorithms this technique can be extended to couniiiig the frequency of execution for active blocks of code. The approaches to the development of update algorithms and the use of active blocksin comparisons will be illustrated by the problem of updating ihe distance matrix of an undirected graph after an edge or vertex has been added or deleted.