Multi-stage modeling using fuzzy multi-criteria feature selection to improve survival prediction of ICU septic shock patients

In many binary medical classification problems, the cost of misclassifying one category is higher than the other, and in these applications it is desirable to employ a classifier with selective sensitivity or specificity. This work explores the utility of a fuzzy multi–criteria function for performance evaluation during knowledge–based medical classification and prediction. The method presented here uses fuzzy optimization to combine the sensitivity, specificity, and accuracy of classification as goals in a single objective function. This approach is used to assign flexible goals, which can be used to maximize the outcome in terms of each one of the goals. The proposed approach significantly increases the sensitivity and the specificity while maintaining or increasing accuracy. The versatility of the method is further exploited in a multi-model approach, using individual structures of multi-objective optimization of sensitivity and specificity separately, and then combining their outcomes through a decision-making module. Among various medical benefits derived from applying this technique, the divergent feature sets selected by high sensitivity and specificity models lend insight into factors more integrally connected to what causes risk of death for patients.     

Assessing the Novelty of Computer-Generated Narratives Using Empirical Metrics

Novelty is a key concept to understand creativity. Evaluating a piece of artwork or other creation in terms of novelty requires comparisons to other works and considerations about the elements that have been reused in the creative process. Human beings perform this analysis intuitively, but in order to simulate it using computers, the objects to be compared and the similarity metrics to be used should be formalized and explicitly implemented. In this paper we present a study on relevant elements for the assessment of novelty in computer-generated narratives. We focus on the domain of folk-tales, working with simple plots and basic narrative elements: events, characters, props and scenarios. Based on the empirical results of this study we propose a set of computational metrics for the automatic assessment of novelty. Although oriented to the implementation of our own story generation system, the measurement methodology we propose can be easily generalized to other creative systems.     

Direct nonlinear control design: the virtual reference feedback tuning (VRFT) approach

This paper introduces the virtual reference feedback tuning (VRFT) approach for controller tuning in a nonlinear setup. VRFT is a data-based method that permits to directly select the controller based on data, with no need for a model of the plant. It is based on a global model reference optimization procedure and, therefore, does not require to access the plant for experiments many times so as to estimate the control cost gradient. For this reason, it represents a very appealing controller design methodology for many control applications.     

A Matrix-Based Approach to the Image Moment Problem

An image can be seen as an element of a vector space and hence it can be expressed in as a linear combination of the elements of any non necessarily orthogonal basis of this space. After giving a matrix formulation of this well-known fact, this paper presents a reconstruction method of an image from its moments that sheds new light on this inverse problem. Two main contributions are presented: (a) the results using the standard approach based on the least squares approximation of the result using orthogonal polynomials can also be obtained using matrix pseudoinverses, which implies higher control on the numerical stability of the problem; and (b) it is possible to use basis functions in the reconstruction different from orthogonal polynomials, such as Fourier or Haar basis, allowing to introduce constraints relative to the bandwidth or the spatial resolution on the image to be reconstructed. Judit Martònez received the B.Sc. degree in 1993 and the PhD degree (with honors) in 1998, both in telecommunications engineering from the Technical University of Catalonia. She developed her research at the Institut de Robútica i Informütica Industrial of the Spanish High Council for Scientific Research. In 1999 she joined the Computer Vision Center, a R&D center founded by the Autonomous University of Barcelona and the Autonomous Government of Catalonia. She has been principal researcher of several industrial and research projects related to computer vision technologies. Her research interests include industrial applications of machine vision, efficient algorithms for low-level image processing, multiresolution mathematical models, statistical clustering, pattern classification and inverse problems. Josep M. Porta received the Engineer Degree in Computer Science in 1994 and the Ph.D. in Artificial Intelligence in 2001, both from the Technical University of Catalonia (UPC). After that, he joined the IAS group of the University of Amsterdam and currently, he holds a post-doc position at the Institut de Robútica i Informütica Industrial (CSIC-UPC) in Barcelona. He carried research in legged robots, machine learning, vision-based methods for autonomous robot localization, and computational kinematics. Federico Thomas is Research Professor at the Spanish Scientific Research Council (CSIC) and director of the Institut de Robútica i Informütica Industrial (CSIC-UPC), Barcelona, Spain. He received the telecommunications engineering degree in 1984, and the Ph.D. degree (with honors) in computer science in 1988, both from the Technical University of Catalonia (UPC). In 1991, he won a NATO postdoctoral scholarship at the University of Massachusetts with the late Prof. Robin Popplestone. In 1999, he was visiting professor, sponsored by the Autonomous Government of Catalonia, at the Oxford University Computing Laboratory with Prof. Stephen Cameron. He has been project leader of several national projects financed by the Spanish Committee for Science and Technology (CICYT), and by local companies such as ENHER, a power generation company now part of ENDESA. His current research interests are in Geometry and Kinematics with applications to Robotics, Computer Graphics and Computer Vision. Prof. Thomas is an Associate Editor of the IEEE Transactions on Robotics.     

Non-Gaussian Model for Ringing Phenomena in Offshore Structures

Significant interest has been shown in identifying the nonlinear mechanisms that induce a ringing type response in offshore structural systems. This high frequency transient type response has been observed in offshore systems, particularly in tension leg platforms (TLPs). Given the implications of this behavior on the fatigue life of TLP tendons, it is essential that ringing be considered in the overall response evaluation. This study presents two non-Gaussian probabilistic models of nonlinear viscous hydrodynamic wave forces that induce ringing. The response of a single-degree-of-freedom system exposed to these non-Gaussian wave force models is then evaluated using analytical and numerical studies based on the It? differentiation rule and the Monte Carlo simulation procedure, respectively. The results demonstrate that the proposed models induce ringing type response in a simplified structure. This study provides a probabilistic framework for modeling ringing type phenomenon which will serve as a building block for more refined hydrodynamic load models.     

Progress in control and data acquisition for the ITER neutral beam test facility

SPIDER, the ion source test bed in the ITER neutral beam test facility, is under construction and its operation is expected to start in 2014. Control and data acquisition for SPIDER are undergoing final design. SPIDER CODAS, as the control and data acquisition system is referred to, is requested to manage 25 plant units, to acquire 1000 analogue signals with sampling rates ranging from a few S/s to 10 MS/s, to acquire images with up to 100 frames per second, to operate with long pulses lasting up to 1 h, and to sustain 200 MB/s data throughput into the data archive with an annual data storage amount of up to 50 TB. SPIDER CODAS software architecture integrates three open-source software frameworks each addressing specific system requirements. Slow control exploits the synergy among EPICS and Siemens S7 programmable controllers. Data handling is by MDSplus a data-centric framework that is geared towards the collection and organization of scientific data. Diagnostics based on imaging drive the design of data throughput and archive size. Fast control is implemented by using MARTe, a data-driven, object-oriented, real-time environment. The paper will describe in detail the progress of the system hardware and software architecture and will show how the software frameworks interact to provide the functions requested by SPIDER CODAS. The paper will focus on how the performance requirements can be met with the described SPIDER CODAS architecture, describing the progress achieved by carrying out prototyping activities.     

An automatic procedure for the synthesis and optimization of very low-frequency Gm–C integrators

Fully integrated low frequency filters are critical cells that should be carefully designed in order to avoid excessive area occupation. In this work we propose an automatic procedure capable of optimizing the design of G_m?CC integrators, which constitute the basis of a wide class of G_m?CC filters. The optimization target is minimizing the cell area with constraints on input range and low frequency noise. Lower and upper bounds can be fixed to most quantities and design parameters in order to avoid solutions that are not compatible with the physical limitations of the process. The program has been developed within the MATLAB^? platform, exploiting the optimization toolbox. The effect of several important design parameters on the optimization of low frequency integrators has been investigated using the proposed routine. The strong interaction between noise and low frequency constraints has been demonstrated, showing the impressive impact of strict noise specifications on the occupied area. The actual effectiveness of parameters such as the current division factor or approaches such as flicker noise rejection by means of chopper modulation has been investigated. Examples of integrator synthesis, performed using the proposed procedure configured with the parameters of a commercial CMOS process, are presented. The consistence between the characteristics of the cells and the initial specifications has been checked using electrical simulations showing a maximum discrepancy with the initial specifications of nearly 80%. A semi-manual method to refine the synthesized cells and improve the accuracy is proposed.     

Familial lead poisoning from contaminated wine

Three members of a family were found be suffering from lead poisoning of nonindustrial origin, the causative agent being the home-produced wine left to ferment in a glazed earthenware vat. The clinical and neurophysiological features are discussed in the light of similar cases in the literature. Correct diagnosis and early treatment of nonindustrial lead polyneuropathy are often extremely difficult because of the elusive nature of the exposure.     

阿维迪钢厂的技术产品优势

1992年,在克日莫那的阿维迪钢厂新建了一条以生产薄板为基础的板材紧凑式生产线,这是欧洲第一、世界第二条薄板生产线.该生产线主要生产高质量钢和特殊钢.数据表明平均每浇次浇11炉钢,56%以上的产品规格都小于2mm厚,经连轧轧成1250mm宽1mm厚的薄板材.可以生产大部分钢种,例如低、中、高碳钢(0.75%),1.2mm的高强度低合金钢,硼钢,耐候钢,合金钢,包晶钢和双相钢等.明年,准备采取措施在一座115t电炉和单流最大宽度为1300mm的生产线上达到100万t的产量.引进铁素体轧制,发展热轧来生产更薄规格的钢以及多相铁素体-马氏体、铁素体-贝氏体钢.和其它薄板生产技术相比,AST(Arvedi Steel Technology阿维迪厂技术)有如下7项优势薄板液芯压下技术,能使板坯中心晶粒细化、温度分布均匀;在线低速轧制(低拉速),从而使中间坯凸度＜1%;板坯厚度方向上的反向温度分布;感应加热炉精确、灵活的温度控制制度;稳定的终轧温度;根据最终薄带的厚度和温度来预设一个恒定的板坯精轧速度;经过克日莫那炉的无头轧制.优良的产品是经过先进的、控制精确的二次、三次冶金生产的.本文将着重陈述阿维迪钢厂的实践要点.并将叙述一些变换成本、高附加值产品及其经济结果.阿维迪有2条年产250万t的生产线,用来生产宽度接近于1600～1800mm的钢种.实践证明,AST的效益归功于它的极高的拉速以及直接无头轧制的思想.目前,AST将上一条高拉速(7.5m/min)的带有精轧的铸轧生产线.阿维迪的ECR的工艺(无头轧制)能够生产0.7mm厚的所有冷轧能生产的同种规格的产品,该工艺还能够生产IF钢(汽车板)、多相钢、薄带甚至有高延伸率和抗拉强度等特殊性能的钢,用来满足汽车工业的需要.克日莫那的阿维迪钢厂用高生产力、低成本生产各种质量的钢种,因此经济效益很好.阿维迪是这项技术的拥有者,准备以阿维迪的名义来购买掉这项技术,其中包括钢液准备、薄带制作以及薄带精整技术.