A drive-by inspection system via vehicle moving force identification

This paper presents a novel method to carry out monitoring of transport infrastructure such as pavements and bridges through the analysis of vehicle accelerations. An algorithm is developed for the identification of dynamic vehicle-bridge interaction forces using the vehicle response. Moving force identification theory is applied to a vehicle model in order to identify these dynamic forces between the vehicle and the road and/or bridge. A coupled half-car vehicle-bridge interaction model is used in theoretical simulations to test the effectiveness of the approach in identifying the forces. The potential of the method to identify the global bending stiffness of the bridge and to predict the pavement roughness is presented. The method is tested for a range of bridge spans using theoretical simulations and the influences of road roughness and signal noise on the accuracy of the results are investigated.

Assessing the evidence: Student response system versus computer based testing for undertaking multiple choice question assessment in undergraduate nursing education

There is a dearth of evidence focusing on student preferences for computer-based testing versus
testing via student response systems for summative assessment in undergraduate education.
This quantitative study compared the preference and acceptability of computer-based testing
and a student response system for completing multiple choice questions in undergraduate
nursing education. After using both computer-based testing and a student response system to
complete multiple choice questions, 192 first year undergraduate nursing students rated their
preferences and attitudes towards using computer-based testing and a student response system.
Results indicated that seventy four percent felt the student response system was easy to use.
Fifty six percent felt the student response system took more time than the computer-based testing
to become familiar with. Sixty Percent felt computer-based testing was more users friendly.
Seventy Percent of students would prefer to take a multiple choice question summative exam
via computer-based testing, although Fifty percent would be happy to take using student response
system. Results are useful for undergraduate educators in relation to student’s preference
for using computer-based testing or student response system to undertake a summative
multiple choice question exam

Temporal and Georaphical Variations in Grass Pollen Seasons in Areas of Western Europe: an Analysis of Season dates at Sites of the European Pollen Information System.

Geographical and temporal variations in the start dates of grass pollen seasons are described for selected sites of the European Pollen Information Service. Daily average grass pollen counts are derived from Network sites in Finland, the Netherlands, Denmark, United Kingdom, Austria, Italy and Spain, giving a broad longitudinal transect over Western Europe. The study is part of a larger project that also examines annual and regional variations in the severity, timing of the peak and duration of the grass pollen seasons. For several sites, data are available for over twenty years enabling long term trends to be discerned. The analyses show notable contrasts in the progression of the seasons annually with differing lag times occurring between southern and northern sites in various years depending on the weather conditions. The patterns identified provide some insight into geographical differences and temporal trends in the incidence of pollinosis. The paper discusses the main difficulties involved in this type of analysis and notes possibilities for using data from the European Pollen Information service to construct pan European predictive models for pollen seasons.

A biosignal embedded system for physiological computing

Thesis submitted in the fulfilment of the requirements for the Degree of Master in Electronic and Telecomunications Engineering

Benchmarking therapeutic drug monitoring software: A review of available computer tools

Therapeutic drug monitoring (TDM) aims to optimize treatments by individualizing dosage regimens based on the measurement of blood concentrations. Dosage individualization to maintain concentrations within a target range requires pharmacokinetic and clinical capabilities. Bayesian calculations currently represent the gold standard TDM approach but require computation assistance. In recent decades computer programs have been developed to assist clinicians in this assignment. The aim of this survey was to assess and compare computer tools designed to support TDM clinical activities. The literature and the Internet were searched to identify software. All programs were tested on personal computers. Each program was scored against a standardized grid covering pharmacokinetic relevance, user friendliness, computing aspects, interfacing and storage. A weighting factor was applied to each criterion of the grid to account for its relative importance. To assess the robustness of the software, six representative clinical vignettes were processed through each of them. Altogether, 12 software tools were identified, tested and ranked, representing a comprehensive review of the available software. Numbers of drugs handled by the software vary widely (from two to 180), and eight programs offer users the possibility of adding new drug models based on population pharmacokinetic analyses. Bayesian computation to predict dosage adaptation from blood concentration (a posteriori adjustment) is performed by ten tools, while nine are also able to propose a priori dosage regimens, based only on individual patient covariates such as age, sex and bodyweight. Among those applying Bayesian calculation, MM-USC*PACK© uses the non-parametric approach. The top two programs emerging from this benchmark were MwPharm© and TCIWorks. Most other programs evaluated had good potential while being less sophisticated or less user friendly. Programs vary in complexity and might not fit all healthcare settings. Each software tool must therefore be regarded with respect to the individual needs of hospitals or clinicians. Programs should be easy and fast for routine activities, including for non-experienced users. Computer-assisted TDM is gaining growing interest and should further improve, especially in terms of information system interfacing, user friendliness, data storage capability and report generation.

Design, development and assessment of the Java Intelligent Tutoring System

The "Java Intelligent Tutoring System" (JITS) research project focused on designing, constructing, and determining the effectiveness of an Intelligent Tutoring System for beginner Java programming students at the postsecondary level. The participants in this research were students in the School of Applied Computing and Engineering Sciences at Sheridan College. This research involved consistently gathering input from students and instructors using JITS as it developed. The cyclic process involving designing, developing, testing, and refinement was used for the construction of JITS to ensure that it adequately meets the needs of students and instructors. The second objective in this dissertation determined the effectiveness of learning within this environment. The main findings indicate that JITS is a richly interactive ITS that engages students on Java programming problems. JITS is equipped with a sophisticated personalized feedback mechanism that models and supports each student in his/her learning style. The assessment component involved 2 main quantitative experiments to determine the effectiveness of JITS in terms of student performance. In both experiments it was determined that a statistically significant difference was achieved between the control group and the experimental group (i.e., JITS group). The main effect for Test (i.e., pre- and postiest), F( l , 35) == 119.43,p < .001, was qualified by a Test by Group interaction, F( l , 35) == 4.98,p < .05, and a Test by Time interaction, F( l , 35) == 43.82, p < .001. Similar findings were found for the second experiment; Test by Group interaction revealed F( 1 , 92) == 5.36, p < .025. In both experiments the JITS groups outperformed the corresponding control groups at posttest.

Arsenic speciation by ion-exchange chromatography with on- line determination by hydride generation: electronic modification of the D.C. plasma by photon counting

A method using L-cysteine for the determination of arsenous acid (As(III)), arsenic acid (As(V)), monomethylarsonic acid (MMAA), and dimethylarsinic acid (DMAA) by hydride generation was demonstrated. The instrument used was a d.c. plasma atomic emission spectrometer (OCP-AES). Complete recovery was reported for As(III), As(V), and DMAA while 86% recovery was reported for MMAA. Detection limits were determined, as arsenic for the species listed previously, to be 1.2, 0.8, 1.1, and 1.0 ngemL-l, respectively. Precision values, at 50 ngemL-1 arsenic concentration, were f.80/0, 2.50/0, 2.6% and 2.6% relative standard deviation, respectively. The L-cysteine reagent was compared directly with the conventional hydride generation technique which uses a potassium iodide-hydrochloric acid medium. Recoveries using L-cysteine when compared with the conventional method provided the following results: similar recoveries were obtained for As(III), slightly better recoveries were obtained for As(V) and MMAA, and significantly better recoveries for DMAA. In addition, tall and sharp peak shapes were observed for all four species when using L-cysteine. The arsenic speciation method involved separation by ion exchange .. high perfonnance liquid chromatography (HPLC) with on-line hydride generation using the L.. cysteine reagent and measurement byOCP-AES. Total analysis time per sample was 12 min while the time between the start of subsequent runs was approximately 20 min. A binary . gradient elution program, which incorporated the following two eluents: 0.01 and 0.5 mM tri.. sodium citrate both containing 5% methanol (v/v) and both at a pH of approximately 9, was used during the separation by HPLC. Recoveries of the four species which were measured as peak area, and were normalized against As(III), were 880/0, 290/0, and 40% for DMAA, MMAA and As(V), respectively. Resolution factors between adjacent analyte peaks of As(III) and DMAA was 1.1; DMAA and MMAA was 1.3; and MMAA and As(V) was 8.6. During the arsenic speciation study, signals from the d.c. plasma optical system were measured using a new photon-signal integrating device. The_new photon integrator developed and built in this laboratory was based on a previously published design which was further modified to reflect current available hardware. This photon integrator was interfaced to a personal computer through an AID convertor. The .photon integrator has adjustable threshold settings and an adjustable post-gain device.

ReAlM - a system to manipulate relations

Given a heterogeneous relation algebra R, it is well known that the algebra of matrices with coefficient from R is relation algebra with relational sums that is not necessarily finite. When a relational product exists or the point axiom is given, we can represent the relation algebra by concrete binary relations between sets, which means the algebra may be seen as an algebra of Boolean matrices. However, it is not possible to represent every relation algebra. It is well known that the smallest relation algebra that is not representable has only 16 elements. Such an algebra can not be put in a Boolean matrix form.[15] In [15, 16] it was shown that every relation algebra R with relational sums and sub-objects is equivalent to an algebra of matrices over a suitable basis. This basis is given by the integral objects of R, and is, compared to R, much smaller. Aim of my thesis is to develop a system called ReAlM - Relation Algebra Manipulator - that is capable of visualizing computations in arbitrary relation algebras using the matrix approach.

Genetic programming for the RoboCup Rescue Simulation System

The Robocup Rescue Simulation System (RCRSS) is a dynamic system of multi-agent interaction, simulating a large-scale urban disaster scenario. Teams of rescue agents are charged with the tasks of minimizing civilian casualties and infrastructure damage while competing against limitations on time, communication, and awareness. This thesis provides the first known attempt of applying Genetic Programming (GP) to the development of behaviours necessary to perform well in the RCRSS. Specifically, this thesis studies the suitability of GP to evolve the operational behaviours required of each type of rescue agent in the RCRSS. The system developed is evaluated in terms of the consistency with which expected solutions are the target of convergence as well as by comparison to previous competition results. The results indicate that GP is capable of converging to some forms of expected behaviour, but that additional evolution in strategizing behaviours must be performed in order to become competitive. An enhancement to the standard GP algorithm is proposed which is shown to simplify the initial search space allowing evolution to occur much quicker. In addition, two forms of population are employed and compared in terms of their apparent effects on the evolution of control structures for intelligent rescue agents. The first is a single population in which each individual is comprised of three distinct trees for the respective control of three types of agents, the second is a set of three co-evolving subpopulations one for each type of agent. Multiple populations of cooperating individuals appear to achieve higher proficiencies in training, but testing on unseen instances raises the issue of overfitting.

The development of a Flanders Interaction Analysis style instrument for use in educational computer mediated communication research

Please consult the paper edition of this thesis to read. It is available on the 5th Floor of the Library at Call Number: Z 9999 E38 D56 1992

A Type-Preserving Compiler from System F to Typed Assembly Language

L'utilisation des méthodes formelles est de plus en plus courante dans le développement logiciel, et les systèmes de types sont la méthode formelle qui a le plus de succès. L'avancement des méthodes formelles présente de nouveaux défis, ainsi que de nouvelles opportunités. L'un des défis est d'assurer qu'un compilateur préserve la sémantique des programmes, de sorte que les propriétés que l'on garantit à propos de son code source s'appliquent également au code exécutable. Cette thèse présente un compilateur qui traduit un langage fonctionnel d'ordre supérieur avec polymorphisme vers un langage assembleur typé, dont la propriété principale est que la préservation des types est vérifiée de manière automatisée, à l'aide d'annotations de types sur le code du compilateur. Notre compilateur implante les transformations de code essentielles pour un langage fonctionnel d'ordre supérieur, nommément une conversion CPS, une conversion des fermetures et une génération de code. Nous présentons les détails des représentation fortement typées des langages intermédiaires, et les contraintes qu'elles imposent sur l'implantation des transformations de code. Notre objectif est de garantir la préservation des types avec un minimum d'annotations, et sans compromettre les qualités générales de modularité et de lisibilité du code du compilateur. Cet objectif est atteint en grande partie dans le traitement des fonctionnalités de base du langage (les «types simples»), contrairement au traitement du polymorphisme qui demande encore un travail substantiel pour satisfaire la vérification de type.

Comparison of epicardial mapping and noncontact endocardial mapping in dog experiments and computer simulations

La fibrillation auriculaire, l'arythmie la plus fréquente en clinique, affecte 2.3 millions de patients en Amérique du Nord. Pour en étudier les mécanismes et les thérapies potentielles, des modèles animaux de fibrillation auriculaire ont été développés. La cartographie électrique épicardique à haute densité est une technique expérimentale bien établie pour suivre in vivo l'activité des oreillettes en réponse à une stimulation électrique, à du remodelage, à des arythmies ou à une modulation du système nerveux autonome. Dans les régions qui ne sont pas accessibles par cartographie épicardique, la cartographie endocardique sans contact réalisée à l'aide d'un cathéter en forme de ballon pourrait apporter une description plus complète de l'activité auriculaire. Dans cette étude, une expérience chez le chien a été conçue et analysée. Une reconstruction électro-anatomique, une cartographie épicardique (103 électrodes), une cartographie endocardique sans contact (2048 électrodes virtuelles calculées à partir un cathéter en forme de ballon avec 64 canaux) et des enregistrements endocardiques avec contact direct ont été réalisés simultanément. Les systèmes d'enregistrement ont été également simulés dans un modèle mathématique d'une oreillette droite de chien. Dans les simulations et les expériences (après la suppression du nœud atrio-ventriculaire), des cartes d'activation ont été calculées pendant le rythme sinusal. La repolarisation a été évaluée en mesurant l'aire sous l'onde T auriculaire (ATa) qui est un marqueur de gradient de repolarisation. Les résultats montrent un coefficient de corrélation épicardique-endocardique de 0.8 (expérience) and 0.96 (simulation) entre les cartes d'activation, et un coefficient de corrélation de 0.57 (expérience) and 0.92 (simulation) entre les valeurs de ATa. La cartographie endocardique sans contact apparait comme un instrument expérimental utile pour extraire de l'information en dehors des régions couvertes par les plaques d'enregistrement épicardique.

Deep learning of representations and its application to computer vision

L’objectif de cette thèse par articles est de présenter modestement quelques étapes du parcours qui mènera (on espère) à une solution générale du problème de l’intelligence artificielle. Cette thèse contient quatre articles qui présentent chacun une différente nouvelle méthode d’inférence perceptive en utilisant l’apprentissage machine et, plus particulièrement, les réseaux neuronaux profonds. Chacun de ces documents met en évidence l’utilité de sa méthode proposée dans le cadre d’une tâche de vision par ordinateur. Ces méthodes sont applicables dans un contexte plus général, et dans certains cas elles on tété appliquées ailleurs, mais ceci ne sera pas abordé dans le contexte de cette de thèse. Dans le premier article, nous présentons deux nouveaux algorithmes d’inférence variationelle pour le modèle génératif d’images appelé codage parcimonieux “spike- and-slab” (CPSS). Ces méthodes d’inférence plus rapides nous permettent d’utiliser des modèles CPSS de tailles beaucoup plus grandes qu’auparavant. Nous démontrons qu’elles sont meilleures pour extraire des détecteur de caractéristiques quand très peu d’exemples étiquetés sont disponibles pour l’entraînement. Partant d’un modèle CPSS, nous construisons ensuite une architecture profonde, la machine de Boltzmann profonde partiellement dirigée (MBP-PD). Ce modèle a été conçu de manière à simplifier d’entraînement des machines de Boltzmann profondes qui nécessitent normalement une phase de pré-entraînement glouton pour chaque couche. Ce problème est réglé dans une certaine mesure, mais le coût d’inférence dans le nouveau modèle est relativement trop élevé pour permettre de l’utiliser de manière pratique. Dans le deuxième article, nous revenons au problème d’entraînement joint de machines de Boltzmann profondes. Cette fois, au lieu de changer de famille de modèles, nous introduisons un nouveau critère d’entraînement qui donne naissance aux machines de Boltzmann profondes à multiples prédictions (MBP-MP). Les MBP-MP sont entraînables en une seule étape et ont un meilleur taux de succès en classification que les MBP classiques. Elles s’entraînent aussi avec des méthodes variationelles standard au lieu de nécessiter un classificateur discriminant pour obtenir un bon taux de succès en classification. Par contre, un des inconvénients de tels modèles est leur incapacité de générer deséchantillons, mais ceci n’est pas trop grave puisque la performance de classification des machines de Boltzmann profondes n’est plus une priorité étant donné les dernières avancées en apprentissage supervisé. Malgré cela, les MBP-MP demeurent intéressantes parce qu’elles sont capable d’accomplir certaines tâches que des modèles purement supervisés ne peuvent pas faire, telles que celle de classifier des données incomplètes ou encore celle de combler intelligemment l’information manquante dans ces données incomplètes. Le travail présenté dans cette thèse s’est déroulé au milieu d’une période de transformations importantes du domaine de l’apprentissage à réseaux neuronaux profonds qui a été déclenchée par la découverte de l’algorithme de “dropout” par Geoffrey Hinton. Dropout rend possible un entraînement purement supervisé d’architectures de propagation unidirectionnel sans être exposé au danger de sur- entraînement. Le troisième article présenté dans cette thèse introduit une nouvelle fonction d’activation spécialement con ̧cue pour aller avec l’algorithme de Dropout. Cette fonction d’activation, appelée maxout, permet l’utilisation de aggrégation multi-canal dans un contexte d’apprentissage purement supervisé. Nous démontrons comment plusieurs tâches de reconnaissance d’objets sont mieux accomplies par l’utilisation de maxout. Pour terminer, sont présentons un vrai cas d’utilisation dans l’industrie pour la transcription d’adresses de maisons à plusieurs chiffres. En combinant maxout avec une nouvelle sorte de couche de sortie pour des réseaux neuronaux de convolution, nous démontrons qu’il est possible d’atteindre un taux de succès comparable à celui des humains sur un ensemble de données coriace constitué de photos prises par les voitures de Google. Ce système a été déployé avec succès chez Google pour lire environ cent million d’adresses de maisons.

Financial Information Mediation: A Case Study of Standards Integration for Electronic Bill Presentment and Payment Using the COIN Mediation Technology

Each player in the financial industry, each bank, stock exchange, government agency, or insurance company operates its own financial information system or systems. By its very nature, financial information, like the money that it represents, changes hands. Therefore the interoperation of financial information systems is the cornerstone of the financial services they support. E-services frameworks such as web services are an unprecedented opportunity for the flexible interoperation of financial systems. Naturally the critical economic role and the complexity of financial information led to the development of various standards. Yet standards alone are not the panacea: different groups of players use different standards or different interpretations of the same standard. We believe that the solution lies in the convergence of flexible E-services such as web-services and semantically rich meta-data as promised by the semantic Web; then a mediation architecture can be used for the documentation, identification, and resolution of semantic conflicts arising from the interoperation of heterogeneous financial services. In this paper we illustrate the nature of the problem in the Electronic Bill Presentment and Payment (EBPP) industry and the viability of the solution we propose. We describe and analyze the integration of services using four different formats: the IFX, OFX and SWIFT standards, and an example proprietary format. To accomplish this integration we use the COntext INterchange (COIN) framework. The COIN architecture leverages a model of sources and receivers’ contexts in reference to a rich domain model or ontology for the description and resolution of semantic heterogeneity.