The transformation of teachers: an adult learning approach

The study focused on the teacher as an adult learner rather than an instructor. A sample of three hundred and three elementary school teachers completed a two-part Likert survey questionnaire. The instrument was developed by the researcher in an attempt to operationalize Mezirow's Theory of Perspective Transformation. The four sub-scales collected information about teachers as they perceived themselves as adult learners and the way they conceptualize critical SelfReflection, Meaning perspectives and New Insights (Mezirow, 1978, 1981, 1989, 1990) within a framework of Mezirow's concept of Transformative Learning. Survey research methodology was used. Frequency distributions, means, and standard deviation were calculated. Reliability analysis and Pearson 'r' correlations established the internal consistency of items It Cross tabulations to describe differences in responses across demographic valuables were computed. The survey results indicated that teachers perceived themselves as self-directed learners. The findings support the need for a better understanding of the teacher as an adult learner so that teacher inservice programs and teacher supervision and evaluation can provide a viable learning alternative to the existing models used in practice.

The role of leaders in building organizational learning capacity

In a world in which social, economic, and environmental circumstances are continuously evolving and increasingly complex, leaders face the challenging prospect of navigating their organizations through unpredictable operating conditions. Finding a way to tap into the learning capacity of the people who comprise their organizations may be the answer to adapt and to survive. This qualitative research study explored the role of leaders in building this organizational learning capacity. The literature identified three domains of personal, interpersonal, and organizational capacity for learning in an organizational setting. Interviews with three senior leaders who had successfully built learning capacity in their respective organizations revealed four elements of leader commitment: (a) to the process of building learning capacity, (b) to organizational objectives and results, (c) to personal actions and behaviours, and (d) to the people of the organization. Each of the four elements of leader commitment spans the three domains of learning capacity that can guide leaders as they build organizational learning capacity.

Teachers' professional learning within communities of mathematics-for-teaching practice

Ontario bansho is an emergent mathematics instructional strategy used by teachers working within communities of practice that has been deemed to have a transformational effect on teachers' professional learning of mathematics. This study sought to answer the following question: How does teachers' implementation of Ontario bansho within their communities of practice inform their professional learning process concerning mathematics-for-teaching? Two other key questions also guided the study: What processes support teachers' professional learning of content-for-teaching? What conditions support teachers' professional learning of content-for-teaching? The study followed an interpretive phenomenological approach to collect data using a purposive sampling of teachers as participants. The researcher conducted interviews and followed an interpretive approach to data analysis to investigate how teachers construct meaning and create interpretations through their social interactions. The study developed a model of professional learning made up of 3 processes, informing with resources, engaging with students, and visualizing and schematizing in which the participants engaged and 2 conditions, ownership and community that supported the 3 processes. The 3 processes occur in ways that are complex, recursive, nonpredictable, and contextual. This model provides a framework for facilitators and leaders to plan for effective, content-relevant professional learning by placing teachers, students, and their learning at the heart of professional learning.

Creating a Space for Self-Transformation: Factors of Success for Adult Literacy Learners with Specific Learning Difficulties

A significant number of adults in adult literacy programs in Ontario have specific learning difficulties. This study sought to examine the holistic factors that contributed to these learners achieving their goals. Through a case study design, the data revealed that a combination of specific learning methods and strategies, along with particular characteristics of the instructor, participant, and class, and the evidence of self-transformation all seemed to contribute to the participant's success in the program. Instructor-directed teaching and cooperative learning were the main learning methods used in the class. General learning strategies employed were the use of core curriculum and authentic documents, and using phonics, repetition, assistive resources, and using activities that appealed to various learning styles. The instructor had a history of both professional development in the area of learning disabilities as well as experience working with learners who had specific learning difficulties. There also seemed to be a goodness of fit between the participant and the instructor. Several characteristics of the participant seemed to aid in his success: his positive self-esteem, self-advocacy skills, self-determination, self-awareness, and the fact that he enjoyed learning. The size (3-5 people) and type of class (small group) also seemed to have an impact. Finally, evidence that the participant went through a self-transformation seemed to contribute to a positive learner identity. These results have implications for practice, theory, and further research in adult education.

Service Learning in Higher Education: A Road Map

As institutions of higher education struggle to stay relevant, competitive, accessible, and flexible, they are scrambling to attend to a shift in focus for new students. This shift involves experiential learning. The purpose of this major research paper was to examine the existing structures, to seek gaps in the experiential learning programs, and to devise a framework to move forward. The specific focus was on experiential learning at Brock University in the Faculty of Applied Health Sciences. The methodology was underscored with cognitive constructivism and appreciative theory. Data collection involved content analysis steps established by Krippendorff (2004) and Weber (1985). Data analysis involved the four dimensions of reflection designed by LaBoskey, including the purpose, context, content, and procedures. The results developed understandings on the state of formal processes and pathways within service learning. A tool kit was generated that defines service learning and offers an overview of the types of service learning typically employed. The tool kit acts as a reference guide for those interested in implementing experiential learning courses. Importantly, the results also provided 10 key points in experiential learning courses by Emily Allan. A flow chart illustrates the connections among each of the 10 points, and then they are described in full to establish a strategy for the way forward in experiential learning.

Whose Voice Is Present?: An Examination of Power and Privilege in Service-Learning in Ontario Universities

This research paper examines themes of power and privilege that occur within service-learning as described by 3 Ontario universities on their service-learning websites. Due to size and time restrictions, this paper was able to examine only 3 Ontario universities: Brock, Wilfrid Laurier, and Lakehead. The purpose of this study is geared towards service-learning practitioners in order for the universities and students to become more self-aware of their immense place of privilege within the service-learning context. Qualitative narrative analysis research methods were employed in this purposeful sample to examine how each university’s story of service-learning reflected themes of power and privilege. The research found that each university posed a unique narrative of service-learning representing various stakeholders’ voices and presence in different ways on their website. Brock largely focuses on faculty and student voices. Laurier intentionally attempts to include all three stakeholder voices, although still favours students and the university as an audience over the community. Lakehead’s unique program includes a plethora of voices and intends much of their information for the community members, students, and the university. The implications of this research demonstrate that universities have a large amount of power and privilege, which is carried through to the students within the service-learning partnership.

Web2.0, knowledge sharing and privacy in E-learning

Quand le E-learning a émergé il ya 20 ans, cela consistait simplement en un texte affiché sur un écran d'ordinateur, comme un livre. Avec les changements et les progrès dans la technologie, le E-learning a parcouru un long chemin, maintenant offrant un matériel éducatif personnalisé, interactif et riche en contenu. Aujourd'hui, le E-learning se transforme de nouveau. En effet, avec la prolifération des systèmes d'apprentissage électronique et des outils d'édition de contenu éducatif, ainsi que les normes établies, c’est devenu plus facile de partager et de réutiliser le contenu d'apprentissage. En outre, avec le passage à des méthodes d'enseignement centrées sur l'apprenant, en plus de l'effet des techniques et technologies Web2.0, les apprenants ne sont plus seulement les récipiendaires du contenu d'apprentissage, mais peuvent jouer un rôle plus actif dans l'enrichissement de ce contenu. Par ailleurs, avec la quantité d'informations que les systèmes E-learning peuvent accumuler sur les apprenants, et l'impact que cela peut avoir sur leur vie privée, des préoccupations sont soulevées afin de protéger la vie privée des apprenants. Au meilleur de nos connaissances, il n'existe pas de solutions existantes qui prennent en charge les différents problèmes soulevés par ces changements. Dans ce travail, nous abordons ces questions en présentant Cadmus, SHAREK, et le E-learning préservant la vie privée. Plus précisément, Cadmus est une plateforme web, conforme au standard IMS QTI, offrant un cadre et des outils adéquats pour permettre à des tuteurs de créer et partager des questions de tests et des examens. Plus précisément, Cadmus fournit des modules telles que EQRS (Exam Question Recommender System) pour aider les tuteurs à localiser des questions appropriées pour leur examens, ICE (Identification of Conflits in Exams) pour aider à résoudre les conflits entre les questions contenu dans un même examen, et le Topic Tree, conçu pour aider les tuteurs à mieux organiser leurs questions d'examen et à assurer facilement la couverture des différent sujets contenus dans les examens. D'autre part, SHAREK (Sharing REsources and Knowledge) fournit un cadre pour pouvoir profiter du meilleur des deux mondes : la solidité des systèmes E-learning et la flexibilité de PLE (Personal Learning Environment) tout en permettant aux apprenants d'enrichir le contenu d'apprentissage, et les aider à localiser nouvelles ressources d'apprentissage. Plus précisément, SHAREK combine un système recommandation multicritères, ainsi que des techniques et des technologies Web2.0, tels que le RSS et le web social, pour promouvoir de nouvelles ressources d'apprentissage et aider les apprenants à localiser du contenu adapté. Finalement, afin de répondre aux divers besoins de la vie privée dans le E-learning, nous proposons un cadre avec quatre niveaux de vie privée, ainsi que quatre niveaux de traçabilité. De plus, nous présentons ACES (Anonymous Credentials for E-learning Systems), un ensemble de protocoles, basés sur des techniques cryptographiques bien établies, afin d'aider les apprenants à atteindre leur niveau de vie privée désiré.

Observational learning of motor skills : Looking for optimal models

L’observation d’un modèle pratiquant une habileté motrice promeut l’apprentissage de l’habileté en question. Toutefois, peu de chercheurs se sont attardés à étudier les caractéristiques d’un bon modèle et à mettre en évidence les conditions d’observation pouvant optimiser l’apprentissage. Dans les trois études composant cette thèse, nous avons examiné les effets du niveau d’habileté du modèle, de la latéralité du modèle, du point de vue auquel l’observateur est placé, et du mode de présentation de l’information sur l’apprentissage d’une tâche de timing séquentielle composée de quatre segments. Dans la première expérience de la première étude, les participants observaient soit un novice, soit un expert, soit un novice et un expert. Les résultats des tests de rétention et de transfert ont révélé que l’observation d’un novice était moins bénéfique pour l’apprentissage que le fait d’observer un expert ou une combinaison des deux (condition mixte). Par ailleurs, il semblerait que l’observation combinée de modèles novice et expert induise un mouvement plus stable et une meilleure généralisation du timing relatif imposé comparativement aux deux autres conditions. Dans la seconde expérience, nous voulions déterminer si un certain type de performance chez un novice (très variable, avec ou sans amélioration de la performance) dans l’observation d’une condition mixte amenait un meilleur apprentissage de la tâche. Aucune différence significative n’a été observée entre les différents types de modèle novices employés dans l’observation de la condition mixte. Ces résultats suggèrent qu’une observation mixte fournit une représentation précise de ce qu’il faut faire (modèle expert) et que l’apprentissage est d’autant plus amélioré lorsque l’apprenant peut contraster cela avec la performance de modèles ayant moins de succès. Dans notre seconde étude, des participants droitiers devaient observer un modèle à la première ou à la troisième personne. L’observation d’un modèle utilisant la même main préférentielle que soi induit un meilleur apprentissage de la tâche que l’observation d’un modèle dont la dominance latérale est opposée à la sienne, et ce, quel que soit l’angle d’observation. Ce résultat suggère que le réseau d’observation de l’action (AON) est plus sensible à la latéralité du modèle qu’à l’angle de vue de l’observateur. Ainsi, le réseau d’observation de l’action semble lié à des régions sensorimotrices du cerveau qui simulent la programmation motrice comme si le mouvement observé était réalisé par sa propre main dominante. Pour finir, dans la troisième étude, nous nous sommes intéressés à déterminer si le mode de présentation (en direct ou en vidéo) influait sur l’apprentissage par observation et si cet effet est modulé par le point de vue de l’observateur (première ou troisième personne). Pour cela, les participants observaient soit un modèle en direct soit une présentation vidéo du modèle et ceci avec une vue soit à la première soit à la troisième personne. Nos résultats ont révélé que l’observation ne diffère pas significativement selon le type de présentation utilisée ou le point de vue auquel l’observateur est placé. Ces résultats sont contraires aux prédictions découlant des études d’imagerie cérébrale ayant montré une activation plus importante du cortex sensorimoteur lors d’une observation en direct comparée à une observation vidéo et de la première personne comparée à la troisième personne. Dans l’ensemble, nos résultats indiquent que le niveau d’habileté du modèle et sa latéralité sont des déterminants importants de l’apprentissage par observation alors que le point de vue de l’observateur et le moyen de présentation n’ont pas d’effets significatifs sur l’apprentissage d’une tâche motrice. De plus, nos résultats suggèrent que la plus grande activation du réseau d’observation de l’action révélée par les études en imagerie mentale durant l’observation d’une action n’induit pas nécessairement un meilleur apprentissage de la tâche.

Leveraging noisy side information for disentangling of factors of variation in a supervised setting

Ce mémoire est composé de trois articles et présente les résultats de travaux de recherche effectués dans le but d'améliorer les techniques actuelles permettant d'utiliser des données associées à certaines tâches dans le but d'aider à l'entraînement de réseaux de neurones sur une tâche différente. Les deux premiers articles présentent de nouveaux ensembles de données créés pour permettre une meilleure évaluation de ce type de techniques d'apprentissage machine. Le premier article introduit une suite d'ensembles de données pour la tâche de reconnaissance automatique de chiffres écrits à la main. Ces ensembles de données ont été générés à partir d'un ensemble de données déjà existant, MNIST, auquel des nouveaux facteurs de variation ont été ajoutés. Le deuxième article introduit un ensemble de données pour la tâche de reconnaissance automatique d'expressions faciales. Cet ensemble de données est composé d'images de visages qui ont été collectées automatiquement à partir du Web et ensuite étiquetées. Le troisième et dernier article présente deux nouvelles approches, dans le contexte de l'apprentissage multi-tâches, pour tirer avantage de données pour une tâche donnée afin d'améliorer les performances d'un modèle sur une tâche différente. La première approche est une généralisation des neurones Maxout récemment proposées alors que la deuxième consiste en l'application dans un contexte supervisé d'une technique permettant d'inciter des neurones à apprendre des fonctions orthogonales, à l'origine proposée pour utilisation dans un contexte semi-supervisé.

Advances in scaling deep learning algorithms

Les algorithmes d'apprentissage profond forment un nouvel ensemble de méthodes puissantes pour l'apprentissage automatique. L'idée est de combiner des couches de facteurs latents en hierarchies. Cela requiert souvent un coût computationel plus elevé et augmente aussi le nombre de paramètres du modèle. Ainsi, l'utilisation de ces méthodes sur des problèmes à plus grande échelle demande de réduire leur coût et aussi d'améliorer leur régularisation et leur optimization. Cette thèse adresse cette question sur ces trois perspectives. Nous étudions tout d'abord le problème de réduire le coût de certains algorithmes profonds. Nous proposons deux méthodes pour entrainer des machines de Boltzmann restreintes et des auto-encodeurs débruitants sur des distributions sparses à haute dimension. Ceci est important pour l'application de ces algorithmes pour le traitement de langues naturelles. Ces deux méthodes (Dauphin et al., 2011; Dauphin and Bengio, 2013) utilisent l'échantillonage par importance pour échantilloner l'objectif de ces modèles. Nous observons que cela réduit significativement le temps d'entrainement. L'accéleration atteint 2 ordres de magnitude sur plusieurs bancs d'essai. Deuxièmement, nous introduisont un puissant régularisateur pour les méthodes profondes. Les résultats expérimentaux démontrent qu'un bon régularisateur est crucial pour obtenir de bonnes performances avec des gros réseaux (Hinton et al., 2012). Dans Rifai et al. (2011), nous proposons un nouveau régularisateur qui combine l'apprentissage non-supervisé et la propagation de tangente (Simard et al., 1992). Cette méthode exploite des principes géometriques et permit au moment de la publication d'atteindre des résultats à l'état de l'art. Finalement, nous considérons le problème d'optimiser des surfaces non-convexes à haute dimensionalité comme celle des réseaux de neurones. Tradionellement, l'abondance de minimum locaux était considéré comme la principale difficulté dans ces problèmes. Dans Dauphin et al. (2014a) nous argumentons à partir de résultats en statistique physique, de la théorie des matrices aléatoires, de la théorie des réseaux de neurones et à partir de résultats expérimentaux qu'une difficulté plus profonde provient de la prolifération de points-selle. Dans ce papier nous proposons aussi une nouvelle méthode pour l'optimisation non-convexe.

Reinforcement Learning Approaches to Power System Scheduling

One major component of power system operation is generation scheduling. The objective of the work is to develop efficient control strategies to the power scheduling problems through Reinforcement Learning approaches. The three important active power scheduling problems are Unit Commitment, Economic Dispatch and Automatic Generation Control. Numerical solution methods proposed for solution of power scheduling are insufficient in handling large and complex systems. Soft Computing methods like Simulated Annealing, Evolutionary Programming etc., are efficient in handling complex cost functions, but find limitation in handling stochastic data existing in a practical system. Also the learning steps are to be repeated for each load demand which increases the computation time.Reinforcement Learning (RL) is a method of learning through interactions with environment. The main advantage of this approach is it does not require a precise mathematical formulation. It can learn either by interacting with the environment or interacting with a simulation model. Several optimization and control problems have been solved through Reinforcement Learning approach. The application of Reinforcement Learning in the field of Power system has been a few. The objective is to introduce and extend Reinforcement Learning approaches for the active power scheduling problems in an implementable manner. The main objectives can be enumerated as:(i) Evolve Reinforcement Learning based solutions to the Unit Commitment Problem.(ii) Find suitable solution strategies through Reinforcement Learning approach for Economic Dispatch. (iii) Extend the Reinforcement Learning solution to Automatic Generation Control with a different perspective. (iv) Check the suitability of the scheduling solutions to one of the existing power systems.First part of the thesis is concerned with the Reinforcement Learning approach to Unit Commitment problem. Unit Commitment Problem is formulated as a multi stage decision process. Q learning solution is developed to obtain the optimwn commitment schedule. Method of state aggregation is used to formulate an efficient solution considering the minimwn up time I down time constraints. The performance of the algorithms are evaluated for different systems and compared with other stochastic methods like Genetic Algorithm.Second stage of the work is concerned with solving Economic Dispatch problem. A simple and straight forward decision making strategy is first proposed in the Learning Automata algorithm. Then to solve the scheduling task of systems with large number of generating units, the problem is formulated as a multi stage decision making task. The solution obtained is extended in order to incorporate the transmission losses in the system. To make the Reinforcement Learning solution more efficient and to handle continuous state space, a fimction approximation strategy is proposed. The performance of the developed algorithms are tested for several standard test cases. Proposed method is compared with other recent methods like Partition Approach Algorithm, Simulated Annealing etc.As the final step of implementing the active power control loops in power system, Automatic Generation Control is also taken into consideration.Reinforcement Learning has already been applied to solve Automatic Generation Control loop. The RL solution is extended to take up the approach of common frequency for all the interconnected areas, more similar to practical systems. Performance of the RL controller is also compared with that of the conventional integral controller.In order to prove the suitability of the proposed methods to practical systems, second plant ofNeyveli Thennal Power Station (NTPS IT) is taken for case study. The perfonnance of the Reinforcement Learning solution is found to be better than the other existing methods, which provide the promising step towards RL based control schemes for practical power industry.Reinforcement Learning is applied to solve the scheduling problems in the power industry and found to give satisfactory perfonnance. Proposed solution provides a scope for getting more profit as the economic schedule is obtained instantaneously. Since Reinforcement Learning method can take the stochastic cost data obtained time to time from a plant, it gives an implementable method. As a further step, with suitable methods to interface with on line data, economic scheduling can be achieved instantaneously in a generation control center. Also power scheduling of systems with different sources such as hydro, thermal etc. can be looked into and Reinforcement Learning solutions can be achieved.

Statistical Machine Learning Techniques for the Prediction of Learning Disabilities in School-Age Children

Learning Disability (LD) is a general term that describes specific kinds of learning problems. It is a neurological condition that affects a child's brain and impairs his ability to carry out one or many specific tasks. The learning disabled children are neither slow nor mentally retarded. This disorder can make it problematic for a child to learn as quickly or in the same way as some child who isn't affected by a learning disability. An affected child can have normal or above average intelligence. They may have difficulty paying attention, with reading or letter recognition, or with mathematics. It does not mean that children who have learning disabilities are less intelligent. In fact, many children who have learning disabilities are more intelligent than an average child. Learning disabilities vary from child to child. One child with LD may not have the same kind of learning problems as another child with LD. There is no cure for learning disabilities and they are life-long. However, children with LD can be high achievers and can be taught ways to get around the learning disability. In this research work, data mining using machine learning techniques are used to analyze the symptoms of LD, establish interrelationships between them and evaluate the relative importance of these symptoms. To increase the diagnostic accuracy of learning disability prediction, a knowledge based tool based on statistical machine learning or data mining techniques, with high accuracy,according to the knowledge obtained from the clinical information, is proposed. The basic idea of the developed knowledge based tool is to increase the accuracy of the learning disability assessment and reduce the time used for the same. Different statistical machine learning techniques in data mining are used in the study. Identifying the important parameters of LD prediction using the data mining techniques, identifying the hidden relationship between the symptoms of LD and estimating the relative significance of each symptoms of LD are also the parts of the objectives of this research work. The developed tool has many advantages compared to the traditional methods of using check lists in determination of learning disabilities. For improving the performance of various classifiers, we developed some preprocessing methods for the LD prediction system. A new system based on fuzzy and rough set models are also developed for LD prediction. Here also the importance of pre-processing is studied. A Graphical User Interface (GUI) is designed for developing an integrated knowledge based tool for prediction of LD as well as its degree. The designed tool stores the details of the children in the student database and retrieves their LD report as and when required. The present study undoubtedly proves the effectiveness of the tool developed based on various machine learning techniques. It also identifies the important parameters of LD and accurately predicts the learning disability in school age children. This thesis makes several major contributions in technical, general and social areas. The results are found very beneficial to the parents, teachers and the institutions. They are able to diagnose the child’s problem at an early stage and can go for the proper treatments/counseling at the correct time so as to avoid the academic and social losses.

The Informational Complexity of Learning from Examples

This thesis attempts to quantify the amount of information needed to learn certain tasks. The tasks chosen vary from learning functions in a Sobolev space using radial basis function networks to learning grammars in the principles and parameters framework of modern linguistic theory. These problems are analyzed from the perspective of computational learning theory and certain unifying perspectives emerge.

Online Learning of Non-stationary Sequences

We consider an online learning scenario in which the learner can make predictions on the basis of a fixed set of experts. The performance of each expert may change over time in a manner unknown to the learner. We formulate a class of universal learning algorithms for this problem by expressing them as simple Bayesian algorithms operating on models analogous to Hidden Markov Models (HMMs). We derive a new performance bound for such algorithms which is considerably simpler than existing bounds. The bound provides the basis for learning the rate at which the identity of the optimal expert switches over time. We find an analytic expression for the a priori resolution at which we need to learn the rate parameter. We extend our scalar switching-rate result to models of the switching-rate that are governed by a matrix of parameters, i.e. arbitrary homogeneous HMMs. We apply and examine our algorithm in the context of the problem of energy management in wireless networks. We analyze the new results in the framework of Information Theory.

The Personal tutorial guidance sessions in the framework of the teaching-learning field of the Bologna process: University of Minho (first cycle) courses students’ perceptions

The Bologna Process defends the adoption of a higher education in teaching-learning methodologies that – in contraposition to the previous model based on the transmission of knowledge, which for being essentially theoretical, gives the student a passive role in the knowledge construction process – allows a (pro) active, autonomous and practical learning, where the student acquires and develops his competences. The personal tutorial guidance sessions are included in the teaching contact hours. This abstract presents a study about the University of Minho (first cycle) Courses Students’ perceptions of the personal tutorial guidance sessions’ relevance in the scope of the learning-teaching process, so as to confirm if the implementation/implantation of the commonly called tutorial (type) education, as an approach to an active, autonomous and practical learning, is sensed by the learners themselves