CULTURE: A STRESSED SYSTEM

As for fuzzy closed loop control systems composed of fuzzy relation de scribed process and controller this paper proposes an analytical method for the stability of fuzzy closed loop control systems using fuzzy relations and a design method for constructing a stable fuzzy controller As a result the paper develops the theory of fuzzy control systems Examples are given to illustrate how to use the method for the analysis and design of fuzzycontrol systems     

The abstract selection task: New data and an almost comprehensive model.

A complete quantitative account of P. Wason's (1966) abstract selection task is proposed. The account takes the form of a mathematical model. It is assumed that some response patterns are caused by inferential reasoning, whereas other responses reflect cognitive processes that affect each card selection separately and independently of other card selections. The model parameters assess the contributions of different interpretational, inferential, and heuristic factors that jointly determine performance in the selection task. The interpretation of most of the model parameters in terms of these different factors is validated experimentally. This model of the selection task is the first to account for the observed frequencies of all 16 possible response patterns that can arise. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

<Emphasis Type="Italic">Vive les differences?</Emphasis> Developing a profile of European information systems research as a basis for international comparisons

The information systems (IS) community is truly international, yet there is often a sense that different elements of the community have different profiles in terms of their research and publication expectations. This paper contributes to this discussion by developing a profile of European IS research that can be used as a basis for international comparisons. It reflects on European research on IS as presented during the first 10 years of the European Conference on Information Systems (ECIS). Based on an analysis of all papers published in the ECIS proceedings during the period 1993–2002, the paper presents the key characteristics of the ECIS conferences, together with a profile of European IS research activity as presented at ECIS. In particular, it highlights the key references and sources used by researchers presenting papers at ECIS. It articulates the research areas presented at ECIS and explores the claim that European IS draws more on social theories than elsewhere. Its contribution in presenting a profile of European research in the IS field lies in identifying particular characteristics of the European style of research that can be compared to that undertaken in other parts of the world.     

A scalable solution to the nearest neighbor search problem through local-search methods on neighbor graphs

Pattern Analysis and Applications - Nearest neighbor search is a powerful abstraction for data access; however, data indexing is troublesome even for approximate indexes. For intrinsically...     

ICRP PUBLICATION 118: ICRP Statement on Tissue Reactions and Early and Late Effects of Radiation in Normal Tissues and Organs - Threshold Doses for Tissue Reactions in a Radiation Protection Context

This report provides a review of early and late effects of radiation in normal tissues and organs with respect to radiation protection. It was instigated following a recommendation in Publication 103 (ICRP, 2007), and it provides updated estimates of 'practical' threshold doses for tissue injury defined at the level of 1% incidence. Estimates are given for morbidity and mortality endpoints in all organ systems following acute, fractionated, or chronic exposure. The organ systems comprise the haematopoietic, immune, reproductive, circulatory, respiratory, musculoskeletal, endocrine, and nervous systems; the digestive and urinary tracts; the skin; and the eye. Particular attention is paid to circulatory disease and cataracts because of recent evidence of higher incidences of injury than expected after lower doses; hence, threshold doses appear to be lower than previously considered. This is largely because of the increasing incidences with increasing times after exposure. In the context of protection, it is the threshold doses for very long follow-up times that are the most relevant for workers and the public; for example, the atomic bomb survivors with 40-50years of follow-up. Radiotherapy data generally apply for shorter follow-up times because of competing causes of death in cancer patients, and hence the risks of radiation-induced circulatory disease at those earlier times are lower. A variety of biological response modifiers have been used to help reduce late reactions in many tissues. These include antioxidants, radical scavengers, inhibitors of apoptosis, anti-inflammatory drugs, angiotensin-converting enzyme inhibitors, growth factors, and cytokines. In many cases, these give dose modification factors of 1.1-1.2, and in a few cases 1.5-2, indicating the potential for increasing threshold doses in known exposure cases. In contrast, there are agents that enhance radiation responses, notably other cytotoxic agents such as antimetabolites, alkylating agents, anti-angiogenic drugs, and antibiotics, as well as genetic and comorbidity factors. Most tissues show a sparing effect of dose fractionation, so that total doses for a given endpoint are higher if the dose is fractionated rather than when given as a single dose. However, for reactions manifesting very late after low total doses, particularly for cataracts and circulatory disease, it appears that the rate of dose delivery does not modify the low incidence. This implies that the injury in these cases and at these low dose levels is caused by single-hit irreparable-type events. For these two tissues, a threshold dose of 0.5Gy is proposed herein for practical purposes, irrespective of the rate of dose delivery, and future studies may elucidate this judgement further.     

Metacomputing across intercontinental networks

An intercontinental network of supercomputers spanning more than 10 000 miles and running challenging scientific applications was realized at the Supercomputing ’99 (SC99) conference in Portland, OR using PACX-MPI and ATM PVCs. In this paper, we describe how we constructed the heterogeneous cluster of supercomputers, the problems we confronted in terms of multi-architecture and the way several applications handled the specific requirements of a metacomputer.     

Industrial implementation of controller performance analysis technology

In this paper, a controller performance monitoring technology is introduced with a focus on its industrial implementation and applications. The technology to be introduced in this paper is known as Performance Analysis Toolbox and Solutions (PATS). The main components are discussed, which include data collection, closed-loop systems identification, control valve stiction monitoring, univariate control performance monitoring, multivariate control performance monitoring, model predictive control performance monitoring, etc. It is shown, through a thorough industry case study with various components of PATS, that the integration of the technology provides more convincing diagnosis results and facilitates identification of control improvement opportunities. Some components of the developed technology can be downloaded from the website http://www.ualberta.ca/～bhuang/research/research.htm.     

Bias detection and estimation in dynamic data reconciliation

A method for detection and estimation of measurement bias in nonlinear dynamic processes is presented. It employs model-based data reconciliation and requires the examination of the resulting difference between the measured and reconciled values. Since bias is commonly present in process measurements, this technique is an important step toward the ultimate goal of reconciling ‘raw’ process data that may contain bias and gross errors in addition to small random errors. A CSTR example shows that this method does allow for the detection of a single bias in a nonlinear dynamic process whether or not the exact model equations are known.     

π-RED*—A graph reducer for a full-fledged γ-calculus

This paper describes an applicative order graph reducer which under interactive control performs high-level program transformations that are governed by the reduction semantics of a full-fledged untyped γ-calculus. To do so with competitive speed, high-level functional programs are by a pre-processor reversibly converted into graph representations of γ-terms. Its subterms are closed by the abstraction of relatively free variables only to the extent absolutely necessary to avoid the complexity of full β-reductions at run-time. Processing these γ-terms is based on high-level interpretation which exploits the simplicity and efficiency of naive graph pointer substitutions when reducing function calls. Partially or completely reduced graphs are by a post-processor re-converted into high-level programs. Full β-reductions are only employed by the pre- and post-processor to maintain correct binding levels when abstracting free variables from γ-terms and when undoing these abstractions or reducing partial applications left over as closures, respectively.