Thesis (M.C.S.)--Carleton University, 2003. / Includes bibliographical references (p. 111-114). Also available in electronic format on the Internet.

Thesis (Ph.D.)--Albert-Ludwigs-Universität Freiburg im Breisgau, 2006. / Includes bibliographical references (p. 201-221) and index.

Thesis (Ph.D.)--Utrecht University, 2006. / Includes bibliographical references (p. [133]-137).

Thesis (Ph. D.)--Psychology, Georgia Institute of Technology, 2006. / Daniel Spieler, Committee Member ; Dennis Folds, Committee Member ; Arthur Fisk, Committee Chair ; Wendy Rogers, Committee Member ; Eric Schumacher, Committee Member.

Multivariate Gaussians: We provide sample complexity upper bounds for semi-agnostically learning multivariate Gaussians under the constraint of approximate differential privacy. These are the first  finite sample upper bounds for general Gaussians which do not impose restrictions on the parameters of the distribution. Our bounds are near-optimal in the case when the covariance is known to be the identity, and conjectured to be near-optimal in the general case. From a technical standpoint, we provide analytic tools for arguing the existence of global "locally small" covers from local covers of the space. These are exploited using modifications of recent techniques for for differentially private hypothesis selection. Mixtures of Gaussians: We consider the problem of learning mixtures of Gaussians under the constraint of approximate differential privacy. We provide the first sample complexity upper bounds for privately learning mixtures of unbounded axis-aligned (or even unbounded univariate) Gaussians. To prove our results, we design a new technique for privately learning mixture distributions. A class of distributions F is said to be list-decodable if there is an algorithm that, given "heavily corrupted" samples from a distribution f in F, outputs a list of distributions, H, such that one of the distributions in H approximates f. We show that if F is privately list-decodable then we can privately learn mixtures of distributions in F. Finally, we show axis-aligned Gaussian distributions are privately list-decodable, thereby proving mixtures of such distributions are privately learnable. / Thesis / Master of Science (MSc) / Is it possible to estimate an unknown probability distribution given random samples from it? This is a fundamental problem known as distribution learning (or density estimation) that has been studied by statisticians for decades, and in recent years has become a topic of interest for computer scientists. While distribution learning is a mature and well understood problem, in many cases the samples (or data) we observe may consist of sensitive information belonging to individuals and well-known solutions may inadvertently result in the leakage of private information. In this thesis we study distribution learning under the assumption that the data is generated from high-dimensional Gaussians (or their mixtures) with the aim of understanding how many samples an algorithm needs before it can guarantee a good estimate. Furthermore, to protect against leakage of private information, we consider approaches that satisfy differential privacy — the gold standard for modern private data analysis.

Studien utgår från en utbildning kring idrotts- och föreningsliv riktad till SFI-studerande i en svensk stad och syftar till att utforska hur ett utbildningsprojekt kan bidra till inkludering och empowerment för målgruppen. För att skapa förståelse kring ämnet undersöktes forskning som bland annat belyser inkludering i en samhällskontext i relation till idrottsrörelsen. Studiens valda teorier genomsyrar metoden i enlighet med deduktion. Metoden bygger på fokusgruppsintervjuer med sammanlagt tolv personer där en semistrukturerad intervjumall användes och datan som framkom har bearbetats grundligt genom kvalitativ analys. Tolkningen har skett med hjälp av Cultural theory of learning och empowerment samt en modell som problematiserar inkludering och integration i samhället. Datan från intervjuerna delades in i två huvudkategorier och fyra subkategorier för att kunna analyseras och tolkas och resultatet visar att inkluderingsprocessen kan delas upp i konkreta och komplexa utsagor. Resultatet pekar vidare på att deltagarna i studien har upplevt utbildningen som stärkande och den har givit dem trygghet och empowerment, samt hjälpt dem att inkluderas i samhället genom att de ökat sin förståelse och respekt för sin omvärld. Det visar också att deltagarna önskar ett större utrymme att skapa goda relationer och kulturella möten snarare än enkelriktad undervisning om svensk kultur med underförstådda krav på anpassning. Detta skulle stärka deras empowerment i lärandet och i samhället och göra inkluderingsprocessen enklare och mer effektiv. Projektet kan också med fördel i framtiden anknytas till kreativa uttrycksformer, för att tillfredsställa olika behov som yttrar sig inom gruppen som utgör objektet i studien.

Nature of the inquiry: My research inquiry investigated how qualities of personal interaction shape and affect the promotion of research utilisation for quality improvement in the US healthcare sector. The research investigated my own professional practice of consulting, teaching, and research regarding the improvement of healthcare practices and outcomes. Efforts to improve the quality of healthcare services are often difficult to realise and sustain. The quality improvement movement in the USA and elsewhere has not conducted much self-examination of its own processes for sources of these perennially problematic results. Relevance: The quality of healthcare services can be readily understood as having consequences of life or death, wellness or suffering. Healthcare expenditures in the USA are estimated at 16% of GDP and over 9% in the UK. Improving healthcare quality improvement efforts is a matter of profound human and social significance. Approach: The DMan research methodology is a reflexively aware process conducted as a cohort and as small learning groups of researchers during the three-year programme. The research inquiry used the complex responsive process of relating theory of learning as emergent changes of meaning or, equivalently, knowledge. As a social science of qualities, it uses the qualities of human interaction as the unit of analysis. The research utilised an interdisciplinary approach drawing upon: healthcare quality improvement literature; organizational discourse studies; research on strategy as practice; performance management; communications theories; the theory of mindful learning; interpersonal neurobiology; figurational sociology; and American pragmatist philosophy. The methodology employs a mindful reflexivity research strategy related to concepts from mindful learning and social neuroscience literature. Central methods included iterative peer and supervisor debriefing and iterative reflexive narrative practice. Findings: A contribution is made to the healthcare literature by describing how ordinary qualities of social coordination dynamics affect the promoters of healthcare research, not just potential users of research. A contribution is made to professional practice by providing a new perspective from which to analyse the sources of performance challenges prevalent in healthcare quality improvement efforts. The research findings indicate how applications of substantial organisational and social resources to promote research utilisation in the US health sector can be co-opted and dissipated away from ostensive substantive objectives. This occurs by research promoters‟ organizational discourse efforts to favourably shape power relating and other qualities of interaction of improvement initiatives. These efforts restrict the emergence of learning about the promoted changes.

This study describes the attempt to understand the quality of mediation between people within a community of practice. An innovative chefs’ apprenticeship in a dedicated restaurant provides a setting in which to explore what happens when a group of young people are learning to become fully accepted members of a community of practice. The setting, the social enterprise of Fifteen London, is founded on a passionate belief in the learning potential of all individuals, regardless of background. Conducted over a period of five years this ethnographic study tells the stories of the apprentices; the story of the community; and the story of conducting the investigation. A pilot study completed in 2005 revealed that beyond the mediation observed between individuals, apprentices’ felt their experience of the culture of the learning environment had a greater impact. The thesis explores the theoretical implications of these findings. Drawn from a sociocultural perspective, two theoretical frameworks are applied: Mediated Learning Experience (Feuerstein, Miller and Tannenbaum, 1994) concerned with the mediation between people and its effect on human development; and Situated Learning (Lave and Wenger, 1991) concerned with participation in a community of practice. These frameworks were felt to be useful to an analysis which demonstrates that a community of practice can be analysed according to the framework for Mediated Learning Experience. The symbiosis of these two approaches creates a coherent framework for discourse in which to analyse the learning process itself. A description of the community highlights the complexities of learning, and the challenges of attempting to change the course of human development by means of cultural transmission and social enterprise. I conclude that this learning environment serves as a good example of what can be achieved when innovation works hand in hand with moral purpose.

This thesis focuses on the problem of "autonomous agents". It is assumed that such agents want to be in a desired state which can be assessed by the agent itself when it observes the consequences of its own actions. Therefore the feedback from the motor output via the environment to the sensor input is an essential component of such a system. As a consequence an agent is defined in this thesis as a self-referential system which operates within a closed sensor- mot or-sensor feedback loop. The generic situation is that the agent is always prone to unpredictable disturbances which arrive from the outside, i.e. from its environment. These disturbances cause a deviation from the desired state (for example the organism is attacked unexpectedly or the temperature in the environment changes, ...). The simplest mechanism for managing such disturbances in an organism is to employ a reflex loop which essentially establishes reactive behaviour. Reflex loops are directly related to closed loop feedback controllers. Thus, they are robust and they do not need a built-in model of the control situation. However, reflexes have one main disadvantage, namely that they always occur 'too late'; i.e., only after a (for example, unpleasant) reflex eliciting sensor event has occurred. This defines an objective problem for the organism. This thesis provides a solution to this problem which is called Isotropic Sequence Order (ISO-) learning. The problem is solved by correlating the primary reflex and a predictive sensor input: the result is that the system learns the temporal relation between the primary reflex and the earlier sensor input and creates a new predictive reflex. This (new) predictive reflex does not have the disadvantage of the primary reflex, namely of always being too late. As a consequence the agent is able to maintain its desired input-state all the time. In terms of engineering this means that ISO learning solves the inverse controller problem for the reflex, which is mathematically proven in this thesis. Summarising, this means that the organism starts as a reactive system and learning turns the system into a pro-active system. It will be demonstrated by a real robot experiment that ISO learning can successfully learn to solve the classical obstacle avoidance task without external intervention (like rewards). In this experiment the robot has to correlate a reflex (retraction after collision) with signals of range finders (turn before the collision). After successful learning the robot generates a turning reaction before it bumps into an obstacle. Additionally it will be shown that the learning goal of 'reflex avoidance' can also, paradoxically, be used to solve an attraction task.

Obesity within the Canadian population is on the rise, as people continue to lead hurried lifestyles there is little time to slow down and prepare a meal for oneself or one’s family. As younger generations grow up in these rushed lifestyles, the opportunities to learn about food and nutrition in the home is quickly disappearing. As a result this practicum project explores promoting healthy lifestyles by reconnecting people with food through the creation of a Culinary Learning Centre located in The Forks Market Building, in Winnipeg, Manitoba. Individuals who are motivated to change the way they view and value food are provided with a comprehensive learning environment that approaches teaching through a holistic and collaborative manner. Informed by an extensive literature review into the Slow Food Movement, and learning theory; research into four case studies; and detailed programming, culminated in an interior design solution promoting positive learning, social interaction, and student well-being. / October 2016