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.

Insights from youth workers regarding young people's engagement in out-of-school-time youth development programs

L’objectif de cette étude exploratoire était de mieux comprendre, du point de vue des animateurs/animatrices qui oeuvrent auprès des jeunes, le phénomène de l’engagement des jeunes dans les programmes parascolaires. Des entretiens qualitatifs ont été entrepris à cette fin avec six animateur(trices). Plusieurs obstacles à l’engagement ont été identifiés, la plupart étant extérieurs aux jeunes. Quelques actions et attitudes spécifiques des animateur(trices) ont été relevées comme importantes pour aider les jeunes à surmonter ces obstacles. De plus, la structure et le contenu du programme, ainsi que l’environnement de groupe, jouent des rôles primordiaux. Les dimensions soulevées par les animateur(trices) ont permis d’enrichir et rendre plus holistique un modèle d’engagement qui a été développé à partir d’une recension des écrits dans ce domaine. Finalement, les implications pour la formation des animateurs(trices) et pour la structure des programmes pour les jeunes ont été explorées.

Beyond Doing Good: An Interview with Dr. Kirsten Johnson on the Canadian Disaster and Humanitarian Response Training Program

Compte-rendu / Review

Marine Engineering Training in India -A Pedagogic Introspection

Indian marine engineers are renowned for employment globally due to their knowledge, skill and reliability. This praiseworthy status has been achieved mainly due to the systematic training imparted to marine engineering cadets. However, in an era of advancing technology, marine engineering training has to remain dynamic to imbibe latest technology as well as to meet the demands of the shipping industry. New subjects of studies have to be included in the curriculum in a timely manner taking into consideration the industry requirements and best practices in shipping. Technical competence of marine engineers also has to be subjected to changes depending upon the needs of the ever growing and over regulated shipping industry. Besides. certain soft skills are to be developed and improved amongst the marine engineers in order to alter or amend the personality traits leading to their career success.If timely corrective action is taken. Indian marine engineers can be in still greater demand for employment in global maritime field. In order to enhance the employability of our mmine engineers by improving their quality, a study of marine engineers in general and class IV marine engineers in particular was conducted based on three distinct surveys, viz., survey among senior marine engineers, survey among employers of marine engineers and survey of class IV marine engineers themselves.The surveys have been planned and questionnaires have been designed to focus the study of marine engineer officer class IV from the point of view of the three distinct groups of maritime personnels. As a result of this, the strength and weakness of class IV marine engineers are identified with regard to their performance on board ships, acquisition of necessary technical skills. employability and career success. The criteria of essential qualities of a marine engineer are classified as academic, technical, social, psychological. physical, mental, emergency responsive, communicative and leadership, and have been assessed for a practicing marine engineer by statistical analysis of data collected from surveys. These are assessed for class IV marine engineers from the point of view of senior marine engineers and employers separately. The Endings are delineated and graphically depicted in this thesis.Besides. six pertinent personality traits of a marine engineer viz. self esteem. learning style. decision making. motivation. team work and listening self inventory have been subjected to study and their correlation with career success have been established wherever possible. This is carried out to develop a theoretical framework to understand what leads a marine engineer to his career attainment. This enables the author to estimate the personality strengths and weaknesses of a serving marine engineer and eventually to deduce possible corrective measures or modifications in marine engineering training in India.Maritime training is largely based on International Conventions on Standard of Training. Certification and Watch keeping for Seafarers 1995. its associated Code and Merchant Shipping (STCW for Seafarers) Rules 1998. Further, Maritime Education, Training and Assessment (META) Manual was subjected to a critical scrutiny and relevant Endings of thc surveys arc superimposed on the existing rule requirement and curriculum. Views of senior marine engineers and executives of various shipping companies are taken into account before arriving at the revision of syllabus of marine engineering courses. Modifications in the pattern of workshop and sea service for graduate mechanical engineering trainees are recommended. Desirable age brackets of junior engineers and chief engineers. use of Training and Assessment Record book (TAR Book) during training etc. have also been evaluated.As a result of the pedagogic introspection of the existing system of marine engineering training in India. in this thesis, a revised pattern of workshop training of six months duration for graduate mechanical engineers. revised pattern of sea service training of one year duration and modified now diagram incorporating the above have been arrived at. Effects of various personality traits on career success have been established along with certain findings for improvement of desirable personality traits of marine engineers.

Farmers’ knowledge and perceptions of potato pests and their management in Uganda

As we initiate entomological research on potato (Solanum tuberosum L.) in Uganda, there is need to understand farmers’ knowledge of existing insect pest problems and their management practices. Such information is important for designing a suitable intervention and successful integrated pest management (IPM) strategy. A farm household survey using a structured questionnaire was conducted among 204 potato farmers in six districts of Uganda (i.e., Kabale, Kisoro, Mbale, Kapchorwa, Mubende, and Kyegegwa) during August and September 2013. Diseases, insect pests, price fluctuations, and low market prices were the four highest ranked constraints in potato production, in order of decreasing importance. Cutworms (Agrotis spp.), aphids (Myzus persicae (Sulzer)), and potato tuber moth (Phthorimaea operculella (Zeller)) were the three most severe insect pests. Ants (Dorylis orantalis Westwood), whiteflies (Bemisia tabaci (Gennadius)), and leafminer flies (Liriomyza huidobrensis (Blanchard)) were pests of moderate importance. Major yield losses are predominantly due to late blight (Phytophthora infestans (Mont.) de Bary) and reached 100% without chemical control in the districts of Kabale, Kisoro, Mbale, and Kapchorwa. On average, farmers had little to moderate knowledge about pest characteristics. The predominant control methods were use of fungicides (72% of respondents) and insecticides (62% of respondents). On average, only 5% of the 204 farmers knew about insect pests and their natural enemies. This lack of knowledge calls for training of both farmers and extension workers in insect pest identification, their biology, and control. Empowering farmers with knowledge about insect pests is essential for the reduction of pesticide misuse and uptake of more environmentally friendly approaches like IPM. Field surveys would need follow-up in order to assess the actual field infestation rates and intensities of each insect pest and compare the results with the responses received from farmers.

How can be strategically improved the international attractivity of the touristic sector in Colombia?

Durante el desarrollo de este trabajo de grado, se describe la posibilidad de incrementar estratégicamente el potencial de atraer visitantes extranjeros al país. Colombia es un lugar lleno de biodiversidad, cultura y belleza; pero algunas circunstancias como la guerra, la mala imagen a nivel internacional y la pobre infraestructura turística, han detenido el crecimiento de los indicadores turísticos. La implementación de diferentes estrategias distintivas para posicionar la industria turística y para lograr el mejoramiento continuo, como la definición de planes de contingencia para luchar en contra de la violencia y la inseguridad, el mejoramiento de las condiciones de infraestructura, la mejor capacitación y entrenamiento de todas las personas relacionadas con este segmento de mercado y también los esfuerzos por difundir en mayor proporción información clara y eficiente sobre Colombia; son algunos de los principales objetivos que deben ser alcanzados para incrementar el nivel de atractividad del sector para visitantes internacionales, y también, estas estrategias contribuirán al mejoramiento de los indicadores de desempeño del sector y del país en general.

Community Health Managers: A Pilot Pedagogic Experience in the University of Rosario

The “Grupo de Estudios en Sistemas Tradicionales de Salud” from the School of Medicine of Universidad del Rosario, in agreement with the “Instituto de Etnobiología”, has designed a training course for a new health agent (the community health manager) meant to consider in its curriculum the difficulties, deficiencies and successes of the Primary Health Care Program. In particular, we have attended OMS suggestions in terms of adequate training of local leaders who should look for self-responsibility and selfdetermination in health care coverage. This training proposal is meant to take into account diverse cultures and traditions in order to offer health care models able to consider cultural particularities, epidemiological profiles, and contextual possibilities, with an intercultural point of view. Hence, the training course’s objective is to offer working tools so that community leaders be able to value and promote traditional health knowledge and practices; seek for food security by means of recovery of traditional productive systems or adaptation of appropriate technologies; environment conservation; use of medicinal plants especially in self-care, and stimulation of community and institutional health promotion activities. Preliminary evaluation suggests that this new health agent will be able to set bridges between communities and health care offers available, always looking for healthy ways of life, culturally and environmentally friendly.

Continuos and intervalic training on cardiovascular fitness in sedentary women

The purpose of this investigation was to evaluate if the intervallic exercise near to the maximum of intensity in a similar fitness population and in women is more effective to improve aerobic capacity than the continuous one of smaller intensity. In order to do it, ten healthy women between 18 and 25 years habitual inhabitants of Bogotá city (located at 2600masl) were selected. They signed the informed consent and were divided randomly in two groups of five participants. They were undergoing to two traditional types of training of ten weeks, three times per week, one hour of duration every day. The continuous training group (Group 1) had a VO2peak intensity of 60%; for his part, the intervallic training group (Group 2) had five sessions of ten minutes of exercise at 70% of VO2peak and 90 seconds of recovery between each session. It was possible to demonstrate that, although the Cardiac Frequency (CF) average of group 2 participants during the exercise sessions were always over those of group 1, this difference was not statistically significant. In the same way, the basal CF and the recovery CF, as well as the VO2peak, did not have significant differences between both groups. The systolic and diastolic blood pressure average, as well as the Respiratory Frequency (FR) were always superior in group 2, and these differences were really significant. It was possible to conclude that there is not a significant difference in cardiovascular fitness between sedentary young women, after 10 weeks of two traditional types of aerobic training, one 60% continuous and other one 70% intervalic.

Description and analysis of pulmonary rehabilitation care programs current state in four cities of Colombia

Purpose: there are many studies reporting the benefits of pulmonary rehabilitation, but few of them exhibit the behavior and activities of these services. This article presents the characteristics of services, parts management and training level of team members, in addition to the variables or instruments used to measure the effectiveness and impact in these programs. Method: it was made a cross sectional convenience sample which included seven pulmonary rehabilitation services in four Colombian cities (Bogotá, Medellín, Manizales and Cali), selected by the coverage, for having at least one year of experience and for being formally established and recognized nationwide. The interdisciplinary team of each service answered a survey that was validated through a pilot test and expert consensus. Participation was voluntary. Results: labor onset pulmonary rehabilitation services correspond to an average of a decade, with COPD and asthma pathologies of attention. The programs are characterized by an outpatient treatment with an average duration of eight to twelve weeks, with a frequency of an hour three times a week. Also, the director of the service is regularly a pulmonologist and the coordinator a physiotherapist (57.14%). The posgradual training of these professionals is notable, and they report to have procedural, administrative and communicative skills, but qualify regular there research skills. The physical and technological resources are well tested. 71.42% have done impact studies, but only 28.57% have been published. All have in common training in upper limbs, lower limbs, respiratory muscles, counseling, functional assessment and quality of life. The effectiveness and impact of programs is measured by the walking test, quality of life questionnaires and activities of daily living.

Rhythmical training for deaf children

This paper examines rhythmic training of hearing impaired children and the relationship between a child’s rhythmical development and his/her speech, short-term memory, sound perception, and physical movement.

Promoting the health of marginalized populations in Ecuador through international collaboration and educational innovations

This paper examines two innovative educational initiatives for the Ecuadorian public health workforce: a Canadian-funded Masters programme in ecosystem approaches to health that focuses on building capacity to manage environmental health risks sustainably; and the training of Ecuadorians at the Latin American School of Medicine in Cuba (known as Escuela Latinoamericana de Medicina in Spanish). We apply a typology for analysing how training programmes address the needs of marginalized populations and build capacity for addressing health determinants. We highlight some ways we can learn from such training programmes with particular regard to lessons, barriers and opportunities for their sustainability at the local, national and international levels and for pursuing similar initiatives in other countries and contexts. We conclude that educational efforts focused on the challenges of marginalization and the determinants of health require explicit attention not only to the knowledge, attitudes and skills of graduates but also on effectively engaging the health settings and systems that will reinforce the establishment and retention of capacity in low- and middle-income settings where this is most needed.

The Acute Care Undergraduate TEaching (ACUTE) initiative: consensus development of core competencies in acute care for undergraduates in the United Kingdom

Background: The care of the acutely ill patient in hospital is often sub-optimal. Poor recognition of critical illness combined with a lack of knowledge, failure to appreciate the clinical urgency of a situation, a lack of supervision, failure to seek advice and poor communication have been identified as contributory factors. At present the training of medical students in these important skills is fragmented. The aim of this study was to use consensus techniques to identify the core competencies in the care of acutely ill or arrested adult patients that medical students should possess at the point of graduation. Design: Healthcare professionals were invited to contribute suggestions for competencies to a website as part of a modified Delphi survey. The competency proposals were grouped into themes and rated by a nominal group comprised of physicians, nurses and students from the UK. The nominal group rated the importance of each competency using a 5-point Likert scale. Results: A total of 359 healthcare professionals contributed 2,629 competency suggestions during the Delphi survey. These were reduced to 88 representative themes covering: airway and oxygenation; breathing and ventilation; circulation; confusion and coma; drugs, therapeutics and protocols; clinical examination; monitoring and investigations; team-working, organisation and communication; patient and societal needs; trauma; equipment; pre-hospital care; infection and inflammation. The nominal group identified 71 essential and 16 optional competencies which students should possess at the point of graduation. Conclusions: We propose these competencies form a core set for undergraduate training in resuscitation and acute care.

Application of neural network to hybrid systems with binary inputs

Boolean input systems are in common used in the electric industry. Power supplies include such systems and the power converter represents these. For instance, in power electronics, the control variable are the switching ON and OFF of components as thyristors or transistors. The purpose of this paper is to use neural network (NN) to control continuous systems with Boolean inputs. This method is based on classification of system variations associated with input configurations. The classical supervised backpropagation algorithm is used to train the networks. The training of the artificial neural network and the control of Boolean input systems are presented. The design procedure of control systems is implemented on a nonlinear system. We apply those results to control an electrical system composed of an induction machine and its power converter.

Animal health challenges and veterinary aspects of organic livestock farming identified through a 3 year EU network project

From 2003-2006, an EU network project ‘Sustaining Animal Health and Food Safety in Organic Farming' (SAFO), was carried out with 26 partners from 20 EU-countries and 4 related partners from 4 candidate or new member states. The focus was the integration of animal health and welfare issues in organic farming with food safety aspects. Four very consistent conclusions became apparent: 1) The climatic, physical and socio-economic conditions vary considerably throughout Europe, leading to different livestock farming systems. This limits the possibility for technology transfer between regions, and creates several challenges for a harmonised regulation, 2) Implementing organic standards at farm level does not always ensure that animal health and welfare reach the high ideals of the organic principles, 3) To overcome these deficiencies, organic farmers and farmer organisations need to take ownership of organic values and, 4) In all participating countries, a strong need for training of farmers and in particular veterinarians in animal health promotion and organic principles was identified. The article presents a summary of papers presented at the five SAFO workshops.

PDM 2006: Proceedings of the International Workshop on Data Mining

Recently major processor manufacturers have announced a dramatic shift in their paradigm to increase computing power over the coming years. Instead of focusing on faster clock speeds and more powerful single core CPUs, the trend clearly goes towards multi core systems. This will also result in a paradigm shift for the development of algorithms for computationally expensive tasks, such as data mining applications. Obviously, work on parallel algorithms is not new per se but concentrated efforts in the many application domains are still missing. Multi-core systems, but also clusters of workstations and even large-scale distributed computing infrastructures provide new opportunities and pose new challenges for the design of parallel and distributed algorithms. Since data mining and machine learning systems rely on high performance computing systems, research on the corresponding algorithms must be on the forefront of parallel algorithm research in order to keep pushing data mining and machine learning applications to be more powerful and, especially for the former, interactive. To bring together researchers and practitioners working in this exciting field, a workshop on parallel data mining was organized as part of PKDD/ECML 2006 (Berlin, Germany). The six contributions selected for the program describe various aspects of data mining and machine learning approaches featuring low to high degrees of parallelism: The first contribution focuses the classic problem of distributed association rule mining and focuses on communication efficiency to improve the state of the art. After this a parallelization technique for speeding up decision tree construction by means of thread-level parallelism for shared memory systems is presented. The next paper discusses the design of a parallel approach for dis- tributed memory systems of the frequent subgraphs mining problem. This approach is based on a hierarchical communication topology to solve issues related to multi-domain computational envi- ronments. The forth paper describes the combined use and the customization of software packages to facilitate a top down parallelism in the tuning of Support Vector Machines (SVM) and the next contribution presents an interesting idea concerning parallel training of Conditional Random Fields (CRFs) and motivates their use in labeling sequential data. The last contribution finally focuses on very efficient feature selection. It describes a parallel algorithm for feature selection from random subsets. Selecting the papers included in this volume would not have been possible without the help of an international Program Committee that has provided detailed reviews for each paper. We would like to also thank Matthew Otey who helped with publicity for the workshop.