Dissertations / Theses on the topic 'Learning process'

Purpose – Due to faster innovation cycles and competitive markets, current methods for implementing and adapting business processes can not keep pace with changing requirements and cause BPM solutions to falls short of business needs. The purpose of this paper is to propose a new approach for implementing an agile BPM methodology by substituting the plan-build-run approach with an incremental prototype-based model, removing intermediaries from the time critical path of business process evolution, and empowering end users to change business processes at runtime by manipulating process artefacts. Design/methodology/approach – Based on interviews with customers and stakeholders and our experience in implementing complex BPM solutions in SMEs, we propose key concepts for an agile BPM approach and derive basic requirements for implementing a BPM system that allows users to redefine business processes during their execution. This analysis is supplemented by a brief overview of current research trends in modelling and implementing agile BPM. Originality/value – All existing solutions examined by our team imply a separate modelling step by users or process managers. The designed key concepts enable users to implicitly model processes without interrupting day to day operations. Our approach enables organisations to introduce business process management in areas where agility is very important (e.g. product development) or to increase operational agility in areas with established BPM. Practical implications – An agile BPM solution can give organisations the flexibility they need to react quickly to changing markets and customer needs. We want to help them to introduce standardization and efficiency without losing agility. In areas where classical BPM is in place, our approach can increase the adaptation rate of process changes. In the areas of knowledge workers with a high level of agility, our approach can increase efficiency by supporting knowledge sharing.

The research site has been the Södra Cell Tofte pulp mill. The main focus in this thesis is how to learn about process control. The need for research on this theme is given implicitly in the foundation and construction of the INPRO programme. Norwegian engineering education is discipline oriented, and the INPRO programme aimed at integrating the three disciplines engineering cybernetics, chemical engineering, and organisation and work life science in a single PhD programme. One goal was to produce knowledge of modern production in chemical process plants based on socio-technical thinking.

In the introduction I outline how my research questions have been developed and the need for doing research in the field of improving process understanding in a continuous process plant.

This thesis provides answer to the three questions:

1. What are the learning systems for workers in a process plant?

2. What is the implication for learning of different socio-technical structures?

3. How can learning be further improved for workers in the process industry?

In order to answer these questions and to provide a background for why these questions are important to Tofte, I describe and analyse the case plant Södra Cell Tofte. I find it necessary to make this part rather extensive in order for the reader to understand the context under which Tofte has been developing its learning arenas or learning systems. I use a sociotechnical framework in doing this. I want to introduce and use this framework as I regard it as useful for one of my purposes with this work: Assisting the production unit at Tofte to improve learning. I go through technological improvements that have been carried out from 1980 onwards, and one major organisational change that has taken place. The downsizing and reorganisation that took place in 1992 is of importance as well as the organisational development effort named “Employeeship” that took place in 1996. I had a leave from the INPRO project for almost a year following and evaluating this particular project. The situation at Tofte in 1994 was lack of good learning systems, and after a major reorganisation in 1992 the organisation defined a need for better responsibility distribution and co-operation.

Chapters 3 and 4 present and discuss theories in order to give a broader background for the research issues in this thesis. In Chapter three I discuss features and characteristics of a continuous process plant as these have consequences on how knowledge and skills can be developed and why process understanding is a necessity. I present socio-technical system thinking (STS) as one way of regarding organisation and management of a process plant, and I further discuss why I find this approach appropriate for providing learning primarily at the shop floor level as an integrated part of daily production.

In Chapter four I argue that knowledge and skills in production are becoming increasingly important in highly automated remotely controlled process plants and develop a theory of “process understanding”. Process understanding is defined as the ability to predict what is going to happen. In order to predict what is going to happen with a system one firstly need to define the system boundaries. This system can then be regarded as a mental model. One must know and analyse input variables (know where and how to get relevant process information), and by this anticipate, like in a mental simulation, what will happen with the parameters within a defined time period. Different possible options may be mentally tested including what will happen if no corrective actions are taken. An ability to predict what is going to happen with the product, to process equipment or any other process variable, has to be developed and refined in order to operate a process plant optimally. Theories state that a variety of knowledge and skills are required and that some skills can be acquired only through years of experience. The “knowing why” within a process plant also has to be strengthen in order to develop a better process understanding, but as an addition to the experience based “knowing how”.

Models of learning regarding the demands given by the production systems in order to develop such process understanding are presented and discussed. These are conventional methods, experiential (problem based) learning, and collective learning. The experiential learning model is discussed and what may inhibit learning from experiences to take place in a plant. I have defined the concept “learning arena” and regarded each shift in a control room as a main learning arena since this is the place where theory meets practice. It is further discussed that practice will differ between shifts within same control room due to different mental models of the process. In various learning arenas, different communities-ofpractice must be joined in order to make more shared mental models with the intention to align different practices.

In Chapter five the research methods used to explore the research questions, and thus to bring forth theories about gaining better process control, are reviewed. I have been inspired by action research methods in order to answer my research questions and to contribute in a necessary change process where development and use of learning arenas have been central. I have been more or less active in these arenas and played back experiences and theories in order to further develop the arenas. Besides participations in learning arenas, methods have been interviews, observations, and written documentations.

Chapter six is the case description of two different kinds of learning systems at Tofte: Operator Training (OT) - Operator based development and execution of education/ training and Operations Workshop (OW) - Problem based learning aiming at better production practice. I have provided background for the two cases as an answer to the educational challenges Tofte had in 1996 and not least to differences in operational practice between shifts. I describe background, characteristics, and development from what I term different learning arenas where the learning about process control will take place. I have also discussed in what ways these two learning systems can be regarded as learning arenas and briefly the kind of learning that can take place in each arena. In two Operations workshops I provide more details in order to show some strengths of the method.

In Chapter seven I provide answers to the three research questions outlined in the introduction and further refined in Chapter 4.5. I analyse how different learning types such as individual, experiential, and collective are covered within different learning arenas and how OT and OW meet the requirements for good learning systems in continuous process plants. Further I analyse how tasks regarding education and training are better distributed in a shift and daytime organisation with the two learning systems, and further how learning is integrated with working and thus process operators’ knowledge and skills are better utilised. When I analyse the implication of learning of two different socio-technical structures, I also regard how managers are better enabled to become facilitators for learning. The two arenas have been well established at Tofte, but to a varying degree in the different departments. When regarding Operator training it is still too early to conclude on its impact on results in the pulp mill. However, two Operations workshops have made positive contributions and demonstrated the potential of the method. The strengths of the methods are the collective learning that place in cogenerated learning arenas. In Operator training this strengthens the master-apprentice method, and in Operations workshops it gives a shared understanding and direction for further tasks in process control. Finally, based on the analysis of the first two questions I discuss how learning for workers can be further improved.

In Chapter eighth I conclude on my theoretical contribution and arguments for further research in the actual fields. Finally, based on my findings I will recommend organisational choices on future actions. STS provides frames and directions for learning to take place within groups along the production line. It is however not the scope of the STS paradigm to provide theories of what constitutes knowledge in operations of a plant. And the STS theories are not developed in order to cover more specifically models for how learning within and across semi-autonomous units and organisational levels may take place. Thus the main contribution of this thesis is learning theories based on two different kinds of learning models as means to develop process understanding.

Constructing internal representations of the world is a fundamental aspect of cognition, allowing us to predict and control our environment. However, sensory observations are frequently noisy and incomplete, leading to the question of how such representations are acquired. We take a normative approach to this problem of inductive inference, or structure learning, asking what agents should believe about the world in light of their observations. We consider experiments from a variety of psychological domains, in each case proposing a rational structure-learning model and comparing real behaviour with model predictions. Firstly, we address the problem of rule learning in a memory-based maze task. We show that gating algorithms, a model-free approach to solving POMDPs, replicate rule acquisition in rats, as well as transfer of learning under rule reversal. Secondly, we consider the problem of behaving flexibly in environments composed of distinct behavioural regimes, or 'contexts'. Vile show that a novel decision-making model that discriminates between contexts captures a number of important animal learning phenomena including spontaneous recovery, partial reinforcement and overlearning effects, and serial reversal learning effects. Thirdly, we turn to perception and whether structure learning can explain participants' behaviour in a perceptual task. Assuming that participants aim to infer the structure underlying observed stimuli, we show that discounting of colour cues when estimating motion direction can be explained by a rational clustering model. Finally, we consider the effect of working memory capacity (WMC) on category learning. Treating category learning as a structure learning problem and modelling WMC as the quantity of inferential resources available, we replicate the positive association between WMC and both rate of learning and ability to switch between categorisation strategies. Our results suggest that the simple assumption that agents construct representations of their environments, combined with adequate modelling tools for representation and inference, can offer parsimonious explanations of behaviour in diverse areas of psychology.

Nonparametric Bayesian inference has widespread applications in statistics and machine learning. In this thesis, we examine the most popular priors used in Bayesian non-parametric inference. The Dirichlet process and its extensions are priors on an infinite-dimensional space. Originally introduced by Ferguson (1983), its conjugacy property allows a tractable posterior inference which has lately given rise to a significant developments in applications related to machine learning. Another yet widespread prior used in nonparametric Bayesian inference is the Beta process and its extensions. It has originally been introduced by Hjort (1990) for applications in survival analysis. It is a prior on the space of cumulative hazard functions and it has recently been widely used as a prior on an infinite dimensional space for latent feature models. Our contribution in this thesis is to collect many diverse groups of nonparametric Bayesian tools and explore algorithms to sample from them. We also explore machinery behind the theory to apply and expose some distinguished features of these procedures. These tools can be used by practitioners in many applications.

Thesis (M.S.)--Massachusetts Institute of Technology, Sloan School of Management, 1997, and Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 1997.
Includes bibliographical references (p. 59-61).
by Timothy H. Ingle.
M.S.

Nowadays training methodology is changing very fast. New modern learning approaches has been developed and integrate in global learning process. All new features of learning oriented and directed into increasing of speed of education process and making level of users knowledge wider. The primary requirements for any learning facility is to provide condition which will be result in forming some new individual knowledge of user and give good mental training to create possibility of user to use all new individual knowledge to find the way to make clear some task which will be faced before users. But the best final study result is to prepare users to give them possibility to work out of rules which means that they should work not in general tasks, but try to make clear more complicated tasks that they don’t have enough ability to do that in fact yet. Of course new knowledge of user of education will be result of learning process (professional level of learning material, learning and teaching tools design, learning process structure facility) and user facilities (mental user characteristic, motivation).

Community renewal is a dominant theme in American society today. It has been said that public administration could and should be a leader in the community renewal movement, yet for the most part the field of public administration fails to â getâ community. This study advances and explores a concept of community learning as part of a broader effort to better understand what a community perspective means for public administration theory and practice. The contributions of this study are two-fold. First, a concept of community learning is drawn from a variety of literature streams that share an ethos of collaborative pragmatism. Community learning occurs when the knowledge created in the integrative â community processâ is fed-forward and embedded at the level of community structure. Furthermore, a â learning communityâ is found where the community learning process is institutionalized at the level of community structure. While community learning is a term being used to some degree in the field of community development, a concept of how communities might learn has yet to be offered. Thus, the conceptualization offered here seeks to fill this gap in the literature. This study also explores the community learning concept empirically in the context of an action research project in Wytheville, Virginia. Here participants worked with a Virginia Tech research team to better understand their community and develop a unified â visionâ for the communityâ s future. The study revealed that the collective or collaborative learning of the â community processâ can occur over time and also in the form of punctuated group â a-haâ moments. In either case, the learning process is one where new knowledge is created in the form of new or altered shared meaning or new ideas. This learning was fed-forward to the community level to become community learning in three ways: 1) as the learning took place in the community field, meaning the participants of the learning process represented the different institutions that make up community structure; 2) through the integrative medium of local media outlets; and 3) through formal and informal processes of knowledge transfer from the group to community level, where the community level was represented by a citizens committee. As communities institutionalize learning processes they can be said to be â learning communities.â Evidence from the Wytheville study provides insights into how this might happen. The implications for the practice of a â new public serviceâ are explored as well as future areas of research relevant to the community learning approach. The study concludes by suggesting what a community perspective for public administration might mean as community learning is a concept based in this perspective.
Ph. D.

This paper is sort of an extended footnote, with a faint Borgesian flavor. What I'm going to do is show how one rather prominent argument in the linguistics literature against one aspect of the research program of behaviorism fails to go through. But I'll also observe that this argument appears to have had no practical effect on linguistic investigations, and that many people seem to assume (tacitly, at least) that this argument fails anyway. So my remarks here don't move the field forward any, but what I hope they do is help to get us all a bit clearer about where we are. The argument I'll be examining, given by Noam Chomsky in Reflections on Language (Chomsky 1975), is against a point of view called "general process learning theory ", a view that regards one goal of psychological theorizing to be the discovery of laws of learning that hold across species and across domains of acquisition. Psychological theorizing is by no means a new development on the linguistics scene. It is true, I think, that in most cases the people who have thought about language (including but not limited to people we would call linguists) have done so against the backdrop of a psychological theory that they assumed to be at least on the right track, and the idea was often to see what you could make of language by applying the analytical tools that the given psychological theory made available. Bloomfield (1926) is an example of this. (For some discussion of Bloomfield's views on psychology, see Lyons 1978, chapter 3.) One also in this context thinks of Piaget, Skinner of course, as well as philosophers of the 17th and 18th centuries of both the continental Cartesian variety and the so-called British Empiricists. I also think it's true that Chomsky's impact on psychology is somewhat unusual in that the flow of influence is in the other direction; that is, the question is, "If human language is like this, then what must the mind be like ?" rather than the other way around. Be that as it may, Chomsky has been, by far and away, the leading expositor of the implications of linguistics for the study of the structure of the human mind. It goes without saying that the ramifications of this work have been very rich, the pivotal role of linguistics in the "cognitive sciences" being just one indication of its influence. One of the earliest engagements at discipline boundaries was Chomsky's forceful assault on B.F. Skinner's attempt to extend the domain of behaviorist psychology to human languages. It's this argument that I want to have another look at. To do this it will be useful to try to isolate several facets of the discussion. I should perhaps reiterate, for the connoisseurs of counterrevolution who I know are out there, that my conclusion will be a modest one. I will not be concluding that after all Skinner was right and Chomsky was wrong. On the contrary, I'm going to assume that this game is over, and has been for quite some time. My goal is to call attention to what I think is an Unsolved problem which acquires its interest because it bears on how we regard linguistics as influencing our judgment about the structure of the human mind.

Abstract The purpose of this thesis is to examine the role of learning and the constituents of learning in creating process innovations and realizing organizational change. Organizational learning was studied and the data was collected in process development projects, in which process simulations were used as a development method for process innovations and change. The empirical research was carried out mainly in 1988–2001. In this thesis, the results of those studies are reflected on, together with the recent literature related to organizational learning theories in the context of process innovations and management of change. Specifically, the concepts of learning enablers, learning outcomes (intangible/tangible), learning paths, and their relationships are studied. Qualitative longitudinal action research with case studies and abductive reasoning are used as the research methods throughout this thesis. The data consists of 34 cases and 99 process simulations in 12 different industries and varies from large core processes to support processes. The case organizations were mainly Finnish companies with Finns as the majority of participants. One case organization was from Switzerland, and in five cases, multicultural groups participated in the process simulations. In each project, process simulations were used at least once during a change project, in some cases even five to eight times. The data analysis proceeded in an abductive manner throughout the included five articles, and the findings are summarized based on the research questions. The theoretical contribution of this thesis is fourfold. The findings give new understanding 1) about learning enablers, their relationships to each other and influence on learning and process innovations, 2) about the role of intangible and tangible learning outcomes in individual and organizational learning processes and 3) about learning paths in process innovations and related change processes. The thesis also defines 4) a model for effective learning processes in change projects concerning process innovations. The contribution in the practical and managerial context should be applicable field of North and West European commercial organizations. This thesis highlights individual and organizational learning in the creation of process innovations, and it also defines the features of an effective development method for creating and implementing process innovations
Tiivistelmä Tämän väitöskirjan tarkoituksena on tutkia organisaation oppimisen ja sen eri osatekijöiden roolia prosessi-innovaatioiden luomisessa ja muutoksen aikaansaamisessa organisaatiossa. Organisaation oppimista tutkittiin ja aineisto kerättiin prosessien kehittämisprojekteissa, joissa prosessisimulaatioita käytettiin prosessi-innovaatioiden menetelmänä. Tutkimuksen aineisto kerättiin pääsiassa vuosien 1988–2001 aikana. Tässä väitöskirjassa näitä tuloksia analysoidaan vallitsevien organisaatio-oppimisen teorioiden valossa prosessi-innovaatioiden ja muutoksen johtamisen asiayhteydessä. Erityisesti tutkitaan oppimisen mahdollistajia, oppimisen tuloksia (aineettomia/aineellisia), oppimisen polkuja ja kaikkien näiden keskinäisiä suhteita. Tutkimusmenetelmänä on käytetty laadullista pitkittäistä toimintatutkimusta tapaustutkimuksineen ja abduktiivisine päättelyineen. Aineisto koostuu 34 tapaustutkimuksesta ja 99 prosessisimulaatiosta, jotka toteutettiin 12 eri toimialalla, ja vaihtelevat ydinprosesseista tukiprosesseihin. Suurimmaksi osaksi tapausorganisaatiot olivat suomalaisia yrityksiä, ja yksi tapausorganisaatio oli sveitsiläinen. Projektiryhmät olivat pääasiassa suomalaisia, mutta monikulttuurinen ryhmä osallistui prosessisimulointiin viidessä tapaustutkimuksessa. Jokaisessa projektissa prosessisimulointia käytettiin ainakin kerran muutosprojektin aikana, joissakin tapauksissa jopa viidestä kahdeksaan kertaan. Väitöskirjan viiden artikkelin aineiston analysointi suoritettiin abduktiivisen päättelyn periaattein, ja tutkimuksen tulokset esitetään tutkimuskysymysten mukaisesti. Tutkimuksen teoreettinen tulos kohdistuu neljään osa-alueeseen. Tulokset lisäävät ymmärrystä 1) oppimisen mahdollistajista, niiden suhteesta toisiinsa ja vaikutuksesta oppimiseen ja prosessi-innovaatioihin, 2) aineettomien ja aineellisten oppimistulosten roolista yksilön ja organisaation oppimisprosesseissa, ja 3) prosessi-innovaatioiden ja niihin liittyvien muutosprosessien oppimisen poluista. Tutkimus myös määrittelee 4) prosessi-innovaatioita koskevien muutosprojektien oppimista korostavan mallin. Tutkimuksen tuloksia voitaneen soveltaa Pohjois- ja Länsi-Euroopan yritysmaailmassa. Tutkimus korostaa yksilön ja organisaation oppimisen merkitystä prosessi-innovaatioiden luomisessa ja lisäksi se määrittelee vaikuttavan prosessi-innovaatioiden kehittämisen ja käyttöönoton menetelmän piirteet

Abstract Knowledge discovery (KD) technology is used to extract knowledge from large quantities of digital data in an automated fashion. The established process model represents the KD process in a linear and technology-centered manner, as a sequence of transformations that refine raw data into more and more abstract and distilled representations. Any actual KD process, however, has aspects that are not adequately covered by this model. In particular, some of the most important actors in the process are not technological but human, and the operations associated with these actors are interactive rather than sequential in nature. This thesis proposes an augmentation of the established model that addresses this neglected dimension of the KD process. The proposed process model is composed of three sub-models: a data model, a workflow model, and an architectural model. Each sub-model views the KD process from a different angle: the data model examines the process from the perspective of different states of data and transformations that convert data from one state to another, the workflow model describes the actors of the process and the interactions between them, and the architectural model guides the design of software for the execution of the process. For each of the sub-models, the thesis first defines a set of requirements, then presents the solution designed to satisfy the requirements, and finally, re-examines the requirements to show how they are accounted for by the solution. The principal contribution of the thesis is a broader perspective on the KD process than what is currently the mainstream view. The augmented KD process model proposed by the thesis makes use of the established model, but expands it by gathering data management and knowledge representation, KD workflow and software architecture under a single unified model. Furthermore, the proposed model considers issues that are usually either overlooked or treated as separate from the KD process, such as the philosophical aspect of KD. The thesis also discusses a number of technical solutions to individual sub-problems of the KD process, including two software frameworks and four case-study applications that serve as concrete implementations and illustrations of several key features of the proposed process model
Tiivistelmä Tiedonlouhintateknologialla etsitään automoidusti tietoa suurista määristä digitaalista dataa. Vakiintunut prosessimalli kuvaa tiedonlouhintaprosessia lineaarisesti ja teknologiakeskeisesti sarjana muunnoksia, jotka jalostavat raakadataa yhä abstraktimpiin ja tiivistetympiin esitysmuotoihin. Todellisissa tiedonlouhintaprosesseissa on kuitenkin aina osa-alueita, joita tällainen malli ei kata riittävän hyvin. Erityisesti on huomattava, että eräät prosessin tärkeimmistä toimijoista ovat ihmisiä, eivät teknologiaa, ja että heidän toimintansa prosessissa on luonteeltaan vuorovaikutteista eikä sarjallista. Tässä väitöskirjassa ehdotetaan vakiintuneen mallin täydentämistä siten, että tämä tiedonlouhintaprosessin laiminlyöty ulottuvuus otetaan huomioon. Ehdotettu prosessimalli koostuu kolmesta osamallista, jotka ovat tietomalli, työnkulkumalli ja arkkitehtuurimalli. Kukin osamalli tarkastelee tiedonlouhintaprosessia eri näkökulmasta: tietomallin näkökulma käsittää tiedon eri olomuodot sekä muunnokset olomuotojen välillä, työnkulkumalli kuvaa prosessin toimijat sekä niiden väliset vuorovaikutukset, ja arkkitehtuurimalli ohjaa prosessin suorittamista tukevien ohjelmistojen suunnittelua. Väitöskirjassa määritellään aluksi kullekin osamallille joukko vaatimuksia, minkä jälkeen esitetään vaatimusten täyttämiseksi suunniteltu ratkaisu. Lopuksi palataan tarkastelemaan vaatimuksia ja osoitetaan, kuinka ne on otettu ratkaisussa huomioon. Väitöskirjan pääasiallinen kontribuutio on se, että se avaa tiedonlouhintaprosessiin valtavirran käsityksiä laajemman tarkastelukulman. Väitöskirjan sisältämä täydennetty prosessimalli hyödyntää vakiintunutta mallia, mutta laajentaa sitä kokoamalla tiedonhallinnan ja tietämyksen esittämisen, tiedon louhinnan työnkulun sekä ohjelmistoarkkitehtuurin osatekijöiksi yhdistettyyn malliin. Lisäksi malli kattaa aiheita, joita tavallisesti ei oteta huomioon tai joiden ei katsota kuuluvan osaksi tiedonlouhintaprosessia; tällaisia ovat esimerkiksi tiedon louhintaan liittyvät filosofiset kysymykset. Väitöskirjassa käsitellään myös kahta ohjelmistokehystä ja neljää tapaustutkimuksena esiteltävää sovellusta, jotka edustavat teknisiä ratkaisuja eräisiin yksittäisiin tiedonlouhintaprosessin osaongelmiin. Kehykset ja sovellukset toteuttavat ja havainnollistavat useita ehdotetun prosessimallin merkittävimpiä ominaisuuksia

Published Article
The relationship between the big five personality traits, Extraversion, Agreeableness Neuroticism, Conscientiousness and Openness to Experience and deep and surface approaches to learning forms the basis of this article. The findings of a research study in this milieu will be presented to prove that earlier studies in this field have been upheld, but that an important deviation has occurred on certain levels of personality. A students way of learning implies the type of learning that is taking place. Ultimately we as lecturers want to encourage deep learning as this stimulates retention of information, important in production of students that are ready for employment.

Abstract. Context and aim: While interacting with peers and teachers in a collaborative learning task, students experience socio-emotional challenges and display emotional responses. These responses have two major components: arousal and valence, which influence the learning process and its outcomes. The aim of the study was twofold: first, to explore how group members’ arousal levels vary across different phases of a collaborative learning task; and second, to investigate how case students’ emotional responses are distributed in the arousal-valence space across the phases of the collaborative task. Methods: Twelve 6th graders from a school of Finland participated in a collaborative task, in groups of three students. The task was to build an energy efficient house in three distinct phases: brainstorming, planning, and building. While performing the activity, students wore Empatica E4 wristbands to measure their electrodermal activity (EDA) and were video-recorded with 360° cameras. Arousal levels were calculated in peaks per min (ppm) and classified as low, middle, and high. Emotional valence was classified from video analysis into positive, neutral, and negative. Results: The ranges for arousal levels were established between 26 and 88 ppm. Only two students displayed the same arousal level across the three phases of the experiment. Three students displayed higher arousal at first and then fell in to lower levels. Four students had the opposite experience and three students did not display a pattern. As for the case students, the student leading a poorly collaborating group experienced oscillating levels of arousal, from middle to high, and displayed a mix of negative and positive valence most of the time. The student loafing around experienced all arousal levels and positive valence most of the time. Overall conclusions and relevance: The study allowed to establish measurement thresholds for arousal as a starting point for future studies in collaborative learning and the arousal-valence space provided a quantifiable picture to help teachers understand the importance of emotional responses in classroom during collaborative learning.

The conventional and optimization based controllers have been used in process industries for more than two decades. The application of such controllers on complex systems could be computationally demanding and may require estimation of hidden states. They also require constant tuning, development of a mathematical model (first principle or empirical), design of control law which are tedious. Moreover, they are not adaptive in nature. On the other hand, in the recent years, there has been significant progress in the fields of computer vision and natural language processing that followed the success of deep learning. Human level control has been attained in games and physical tasks by combining deep learning with reinforcement learning. They were also able to learn the complex go game which has states more than number of atoms in the universe. Self-Driving cars, machine translation, speech recognition etc started to gain advantage of these powerful models. The approach to all of them involved problem formulation as a learning problem. Inspired by these applications, in this work we have posed process control problem as a learning problem to build controllers to address the limitations existing in current controllers.
Applied Science, Faculty of
Chemical and Biological Engineering, Department of
Graduate

Performing a set-up on a complex machine may be difficult. This problem arises frequently for the industry, especially when the relation between input and output data cannot be defined precisely. Heavy methods of optimization may be used to perform a set-up. This master thesis investigate the possibility to use a machine learning approach on a specific machine. We study the structure of the relation between input and output data. We show the variations are smooth. We define a set of tests to evaluate future models. We design and test several models on simulation data, and select the best one. We design a strategy to use data in the best possible way. The selected model is then tested on actual data in order to be optimized.
Att justera ett komplext maskineri kan vara mödosamt. Detta problem uppkommer ofta i industrin när relationen mellan in- och utdata inte kan definieras. Tunga optimeringsmetoder kan användas för justeringprocessen. Dennauppsats undersöker möjligheten att använda en s.k. machine learning approach med ett specifikt maskineri. Vi studerar datastrukturen och relationen mellan in- och utdata. Vi visar på att variationer är släta. Vi definierar en rad av tester för att värdera framtida modeller. Vi formger och testar flera modeller med simulerade data och välja den bästa. Vi designar en strategi för att använda data på bästa sätt, för att sedan testa den utvalda modellen på verklig data för att optimeras.

Although it seems clear that coaches become effective practitioners through idiosyncratic combinations of learning experiences (Werthner & Trudel, 2009), little is known about how and why this occurs and impacts on coaching knowledge and practice (Cushion et al., 2010). This research sought to understand the processes and impact of coaches learning in the context of UK youth football coaching, specifically centring on a formal education course. The research process utilised a pragmatic and integrated perspective, influenced by impact evaluation frameworks (e.g. Coldwell & Simkins, 2011). A group of 25 coaches were investigated at different points over a period of a year and a half, using a mixture of semi-structured interviews, systematic observations, video-based stimulated recall interviews and course observations, to build up increasingly in-depth levels of data. Using the principles of grounded theory methodology (Strauss & Corbin, 1998) as well as mixed analyses of variance (ANOVA), changes in the knowledge use and practice behaviours of course candidates, and equivalent coaches not undertaking formal education, were compared. The course had subtle impacts on coaches knowledge conceptions in interaction with wider knowledge sources, yet impact on practice was generally demonstrated only in the areas of questioning content and individually directed coaching interventions. Mismatches between the espoused theory of the course and what the candidates actually perceived, as well as a lack of individualised support to overcome disjuncture (Jarvis, 2006) created barriers to learning, preventing integration of theoretical conceptions into altered coaching practice. A substantive grounded theory was generated to explain the underpinning double-loop cognitive filter and reflective feedback processes involved in coaches learning. The model demonstrated that practitioners learning, guided by existing biography and driven by a practical focus on what works , was heavily influenced by context. Meaningful learning connected knowledge with implementation in practice through reflection. These processes help explain uneven learning across individuals; addressing for the first time questions of what works , how , 'why', and for whom in coach learning (McCullick et al., 2009). Thus the results generate an understanding of coaches learning which can be practically relevant in fostering better opportunities to enhance the development of capable and creative coaches.

Computer software for music has made a significant impact by affecting the perspective of music making, music creating, music education, music production, and music distribution. This impact continues to evolve as individuals seek new avenues of musical expression. Through the papers included in this document, I seek to explore the range and impact of computer software in music, especially software related to music creativity and composition. The first paper is a review of literature concerning the effect of software on creative thought, creativity in music, and the influence this has had in musical composition. In this paper I also explore various techniques of composition, including linear and non-linear processes, which make use of the computer and music software. The second paper is a case study examining the use of music software in a compositional process. In this paper I explore the experiences of the student who was learning to compose music using music software. It offers the perspectives of the students as they developed through each step of the compositional process. It is important that the reader understand the distinction between composition and improvisation as discussed in these two papers. Simply stated, the act of composing is described as the process of forming, making, creating, or constructing music with various elements, phrases, or sections of music. Composing music is often associated with a skill set that captures inspiration and transforms it into a permanent record. Improvisation is the act of creating and playing new music without specific forethought or prior preparation. It can also be define as a skill of creating music in a spontaneous, impromptu, or impulsive way.

Cette thèse porte sur la conception d’outils interactifs pour comprendre et faciliter l’apprentissage de la danse à partir de vidéos. Les processus d’apprentissage des danseurs représentent une source d’informations riches pour les chercheurs qui s’intéressent à la conception de systèmes soutenant l’apprentissage moteur. En effet, les danseurs experts réutilisent un large éventail de compétences qu’ils ont appris. Cependant, ces compétences sont en partie le résultat de connaissances implicites et incarnées,qui sont difficilement exprimables et verbalisables par un individu.Dans cette thèse, je soutiens que nous pouvons capturer et sauvegarder une trace des connaissances implicites des danseurs et les utiliser pour concevoir des outils interactifs qui soutiennent l’apprentissage de la danse. Mon approche consiste à étudier différentes sessions d’apprentissage de danse dans des contextes réels, aussi bien individuels que collaboratifs.Sur la base des résultats apportés par ces études, je contribue à une meilleure compréhension des processus implicites qui sous-tendent l’apprentissage de la danse dans des contextes individuels et collectifs. Je présente plusieurs stratégies d’apprentissage utilisées par des danseurs et j’affirme que l’on peut documenter ces stratégies en sauvegardant une trace de l’apprentissage. Je discute de l’opportunité que représente la capture de ces connaissances incarnées et j’apporte de nouvelles perspectives pour la conception d’outils d’aide à l’apprentissage du mouvement par la vidéo
This thesis focuses on designing interactive tools to understand and support dance learning from videos. Dancers' learning practice represents a rich source of information for researchers interested in designing systems that support motor learning. Indeed, dancers embody a wide range of skills that they reuse during new dance sequences learning. However, these skills are in part the result of embodied implicit knowledge. In this thesis, I argue that we can capture and save traces of dancers' embodied knowledge and use them to design interactive tools that support dance learning. My approach is to study real-life dance learning tasks in individual and collaborative settings. Based on the findings from all the studies, I discuss the challenge of capturing embodied knowledge to support dancers’ learning practice. My thesis highlights that although dancers’ learning processes are diverse, similar strategies emerge to structure their learning process. Finally, I bring and discuss new perspectives to the design of movement-based learning tools

Dissertação (mestrado) - Universidade Federal de Santa Catarina, Centro de Comunicação e Expressão
Made available in DSpace on 2012-10-17T07:42:14Z (GMT). No. of bitstreams: 0Bitstream added on 2016-01-09T00:17:37Z : No. of bitstreams: 1 149142.pdf: 3347623 bytes, checksum: de70e90b779b26ccbf1d0362988ac7cf (MD5)
No contexto desta pesquisa, a tradução é vista como um processo mental que pode estabelecer importantes traços lingüísticos na memória, os quais podem ser utilizados na produção lingüística, no caso desta pesquisa, em resumos escritos. Vinte e sete alunos brasileiros graduandos do curso de inglês, considerados avançados, participaram do estudo. Os resultados forneceram subsídios a uma reavaliação da função da tradução na aquisição da lingua estrangeira. Parece que a tradução tende a levar os alunos a articular mais o sistema lingüístico, o que faz com que eles notem alguns elementos do texto original e os incluam nos resumos. Abstract : The use of translation in second language teaching has been a debatable matter for a long time. In the context of this research, translation is seen as a mental process which may lead to awareness of linguistic features of the input text. By taking into consideration that translating involves an intense elaborate linguistic processing, it was hypothesized that the translation process may estabiish linguistíc traces in memory which may, then, be used in language production, as in the case of this study - summary writing. In order to verify this hypothesis, the summaries produced after the activities of reading and translating were compared with summaries produced after a simple reading comprehension activity. Twenty-seven advanced undergraduate Brazilian students, from the English departments of two Brazilian universities (UFSC and UFRGS), participated in the study. Data were collected in two sessions. In the first session leamers read a text in English, translated into Portuguese and wrote a summary in English (T1). In the second session learners read a text in English and wrote a summary in English as well (T2). In both tasks learners produced summaries without having access to the original text. The results provide ground for a reapraisal of the role of translation in foreign language acquisition. Despite the small differences, the data allows us to conclude that the translation activity produced different results ín the summaries than the reading activity. It appears that translation leads learners to articulate more their language system, and makes them notice some features of the input material and include these features in the summaries. However, more studies are needed before generalizations can be made in relation to the results of this research.

L'apprentissage automatique (ML) a connu d'énormes succès ces dernières années et repose sur un nombre toujours croissant d'applications réelles. Cependant, la conception d'algorithmes prometteurs pour un problème spécifique nécessite toujours un effort humain considérable. L'apprentissage automatique (AutoML) a pour objectif de sortir l'homme de la boucle. AutoML est généralement traité comme un problème de sélection d’algorithme / hyper-paramètre. Les approches existantes incluent l’optimisation Bayésienne, les algorithmes évolutionnistes et l’apprentissage par renforcement. Parmi eux, auto-sklearn, qui intègre des techniques de meta-learning à l'initialisation de la recherche, occupe toujours une place de choix dans les challenges AutoML. Cette observation a orienté mes recherches vers le domaine du meta-learning. Cette orientation m'a amené à développer un nouveau cadre basé sur les processus de décision Markovien (MDP) et l'apprentissage par renforcement (RL). Après une introduction générale (chapitre 1), mon travail de thèse commence par une analyse approfondie des résultats du Challenge AutoML (chapitre 2). Cette analyse a orienté mon travail vers le meta-learning, menant tout d’abord à proposer une formulation d’AutoML en tant que problème de recommandation, puis à formuler une nouvelle conceptualisation du problème en tant que MDP (chapitre 3). Dans le cadre du MDP, le problème consiste à remplir de manière aussi rapide et efficace que possible une matrice S de meta-learning, dans laquelle les lignes correspondent aux tâches et les colonnes aux algorithmes. Un élément de matrice S (i, j) est la performance de l'algorithme j appliqué à la tâche i. La recherche efficace des meilleures valeurs dans S nous permet d’identifier rapidement les algorithmes les mieux adaptés à des tâches données. Dans le chapitre 4, nous examinons d’abord le cadre classique d’optimisation des hyper-paramètres. Au chapitre 5, une première approche de meta-learning est introduite, qui combine des techniques d'apprentissage actif et de filtrage collaboratif pour prédire les valeurs manquantes dans S. Nos dernières recherches appliquent RL au problème du MDP défini pour apprendre une politique efficace d’exploration de S. Nous appelons cette approche REVEAL et proposons une analogie avec une série de jeux pour permettre de visualiser les stratégies des agents pour révéler progressivement les informations. Cette ligne de recherche est développée au chapitre 6. Les principaux résultats de mon projet de thèse sont : 1) Sélection HP / modèle : j'ai exploré la méthode Freeze-Thaw et optimisé l'algorithme pour entrer dans le premier challenge AutoML, obtenant la 3ème place du tour final (chapitre 3). 2) ActivMetaL : j'ai conçu un nouvel algorithme pour le meta-learning actif (ActivMetaL) et l'ai comparé à d'autres méthodes de base sur des données réelles et artificielles. Cette étude a démontré qu'ActiveMetaL est généralement capable de découvrir le meilleur algorithme plus rapidement que les méthodes de base. 3) REVEAL : j'ai développé une nouvelle conceptualisation du meta-learning en tant que processus de décision Markovien et je l'ai intégrée dans le cadre plus général des jeux REVEAL. Avec un stagiaire en master, j'ai développé des agents qui apprennent (avec l'apprentissage par renforcement) à prédire le meilleur algorithme à essayer. Le travail présenté dans ma thèse est de nature empirique. Plusieurs méta-données du monde réel ont été utilisées dans cette recherche. Des méta-données artificielles et semi-artificielles sont également utilisées dans mon travail. Les résultats indiquent que RL est une approche viable de ce problème, bien qu'il reste encore beaucoup à faire pour optimiser les algorithmes et les faire passer à l’échelle aux problèmes de méta-apprentissage plus vastes
Machine Learning (ML) has enjoyed huge successes in recent years and an ever- growing number of real-world applications rely on it. However, designing promising algorithms for a specific problem still requires huge human effort. Automated Machine Learning (AutoML) aims at taking the human out of the loop and develop machines that generate / recommend good algorithms for a given ML tasks. AutoML is usually treated as an algorithm / hyper-parameter selection problems, existing approaches include Bayesian optimization, evolutionary algorithms as well as reinforcement learning. Among them, auto-sklearn which incorporates meta-learning techniques in their search initialization, ranks consistently well in AutoML challenges. This observation oriented my research to the Meta-Learning domain. This direction led me to develop a novel framework based on Markov Decision Processes (MDP) and reinforcement learning (RL).After a general introduction (Chapter 1), my thesis work starts with an in-depth analysis of the results of the AutoML challenge (Chapter 2). This analysis oriented my work towards meta-learning, leading me first to propose a formulation of AutoML as a recommendation problem, and ultimately to formulate a novel conceptualisation of the problem as a MDP (Chapter 3). In the MDP setting, the problem is brought back to filling up, as quickly and efficiently as possible, a meta-learning matrix S, in which lines correspond to ML tasks and columns to ML algorithms. A matrix element S(i, j) is the performance of algorithm j applied to task i. Searching efficiently for the best values in S allows us to identify quickly algorithms best suited to given tasks. In Chapter 4 the classical hyper-parameter optimization framework (HyperOpt) is first reviewed. In Chapter 5 a first meta-learning approach is introduced along the lines of our paper ActivMetaL that combines active learning and collaborative filtering techniques to predict the missing values in S. Our latest research applies RL to the MDP problem we defined to learn an efficient policy to explore S. We call this approach REVEAL and propose an analogy with a series of toy games to help visualize agents’ strategies to reveal information progressively, e.g. masked areas of images to be classified, or ship positions in a battleship game. This line of research is developed in Chapter 6. The main results of my PhD project are: 1) HP / model selection: I have explored the Freeze-Thaw method and optimized the algorithm to enter the first AutoML challenge, achieving 3rd place in the final round (Chapter 3). 2) ActivMetaL: I have designed a new algorithm for active meta-learning (ActivMetaL) and compared it with other baseline methods on real-world and artificial data. This study demonstrated that ActiveMetaL is generally able to discover the best algorithm faster than baseline methods. 3) REVEAL: I developed a new conceptualization of meta-learning as a Markov Decision Process and put it into the more general framework of REVEAL games. With a master student intern, I developed agents that learns (with reinforcement learning) to predict the next best algorithm to be tried. To develop this agent, we used surrogate toy tasks of REVEAL games. We then applied our methods to AutoML problems. The work presented in my thesis is empirical in nature. Several real world meta-datasets were used in this research. Artificial and semi-artificial meta-datasets are also used in my work. The results indicate that RL is a viable approach to this problem, although much work remains to be done to optimize algorithms to make them scale to larger meta-learning problems

A Markov chain is introduced to the major steps of the process of learning a subject matter by a group of students in the classroom, in order to obtain a mathematical representation of the above process. A classroom experiment for learning mathematics is also presented illustrating the applicability of our results in practice.

Abstract Technological advancement in the field of information-and-communication technologies (ICT) was rapid during the first decade of the new millennium. Universities started to use the new information-and-communication technologies more in their core processes, which speeded up their transformation from the traditional campus mode toward virtual universities. Research done in this thesis first investigates the traditional campus university’s change process toward the virtual university model. During the implementation process a geographically distributed e-learning concept was also developed for university use. This concept was transferred and researched in the small and medium enterprise (SME) context in the last part of this research. In large and complex organizations such as universities, it is difficult to find out how the change really was implemented. The literature on change management is voluminous but is dominated by descriptions of single projects. To overcome the limitations of such case studies, this research applies a longer and wider perspective to the change process and, by the introduction of an overarching method that categorizes the investments, shows more clearly the trends, stages of, and the barriers to the development. This long-term study is based on 116 development projects during a ten-year period in a decentralized and networked development environment. In the company setting, conventional training is being replaced more and more by e-learning. To scaffold SMEs in their e-learning adaptation, the concept of e-learning was transferred to the SME environment. The company cases’ part of the thesis presents how the transferability of geographically distributed e-learning concept was developed and tested in the SME environment. As a result, the principles that guided ICT strategy implementation and how the strategies were implemented during a ten-year period are presented. The concept for geographically distributed e-learning environments and its development are also introduced. Finally, the process and results from the concept implementation to the SME environment are presented. This thesis presents university management with an understanding of how larger long-term trends give us the possibility to better understand today’s fast-paced changes. It also gives company managers an example of how models developed in the university environment can be transferred to the company environment
Tiivistelmä Tietotekniikan kehitys on ollut nopeaa vuosituhannen vaihteen jälkeen monilla aloilla. Yliopistoissa sitä on alettu käyttää yhä enemmän osana perustoimintoja, mikä on edistänyt niiden kehitystä traditionaalisista kampusyliopistoista kohti virtuaaliyliopistoja. Tämän tutkimuksen ensimmäisessä osassa on tarkasteltu perinteisen kampusyliopiston muutosprosessia kohti aktiivisesti uutta teknologiaa hyödyntävää yliopistoa ja toisessa osassa muutosten yhtenä tuloksena syntynyt konsepti siirrettiin yritysympäristöön. Yliopistossa muutos toteutettiin määrittelemällä ensin tavoitteet strategioiksi ja sen jälkeen niiden toimeenpanoprosessin avulla. Suurissa organisaatioissa, kuten yliopistoissa, on vaikea hahmottaa kuinka muutokset todellisuudessa toteutetaan. Kirjallisuutta ja tutkimusta organisaatiomuutoksen toteuttamisesta on paljon, mutta suuri osa siitä keskittyy tarkastelemaan yhden projektin aikana tapahtunutta muutosta. Tässä tutkimuksessa on pyritty esittämään laajempi ja pitkäkestoisempi kuva muutosprosessista. Tutkimus pohjautuu 116 kehitysprojektin saamaan rahoitukseen kymmenen vuoden ajalta. Sen avulla on tarkasteltu yliopiston kehitystrendejä ja keskeisiä kehityskohteita. Rahoitus on jaettu projekteille strategioissa määriteltyjen tavoitteiden mukaisesti strategioiden toimeenpanemiseksi. Osana toimeenpanoprosessia syntyi yliopistokäyttöön maantieteellisesti hajautetun oppimisympäristön konsepti. Yritysympäristössä perinteinen koulutus on korvattu yhä useammin teknologiaa hyödyntävillä koulutuksen muodoilla. Pienillä ja keskisuurilla yrityksillä on kuitenkin varsin rajoitetut resurssit koulutuksessa käytettävien teknologioiden käyttöönottoon ja tehokkaaseen käyttöön. Tämän tutkimuksen toisessa osassa yrityksen maantieteellisesti hajautetun oppimisympäristön käyttöönottoa pyrittiin tukemaan tarjoamalla sen käyttöön valmis yliopistoympäristössä kehitetty konsepti. Tämän konseptin siirrettävyyttä ja käyttöönottoa tutkittiin tutkimuksen toisessa osassa. Tämän tutkimuksen tuloksena esitetään periaatteet, jotka ohjasivat yliopiston tieto- ja viestintätekniikan opetuskäytön strategian laadintaa ja toimeenpanoa vuosina 2000–2009. Lisäksi esitellään maantieteellisesti hajautetun oppimisympäristön konsepti. Lopuksi esitellään konseptin käyttöönottoa yritysympäristössä. Tämä väitöskirja esittelee yliopiston johdolle ja toimijoille, kuinka pitkäkestoisten trendien tunteminen auttaa ymmärtämään nykyhetkeä ja sen nopeita muutoksia. Yritysjohdolle väitöskirja tarjoaa esimerkin, kuinka yliopistoympäristöön kehitetty konsepti voidaan siirtää yritysympäristöön

I work within the Higher and Further Education sector, and part of my work focuses on addressing the reasons why students fail to complete level 2 courses (courses equivalent to 5 GCSEs grade A-C). Having qualified as a practising Life Coach, and reflecting upon the knowledge I gained, I became convinced that a life coach intervention course could make the difference between a student dropping out of a course or remaining to course completion. To test my assumption, I undertook a three year Action Research study focusing on a specific level 2 programme (containing the largest cohort of learners within my portfolio remit). The research question started out as: Can life coaching be effectively used as a tool to support learners to course completion? In year one I collected data to establish the views of Employer Training Pilot (ETP) learners, . covering the barriers they faced to ongoing learning and also data from qualified Life Coaches to identify how specific aspects of life coaching could address improving retention and achievement for this cohort. '' ' The outcome of the evaluation culminated in a course (L1FEGIM) that potentially provided the support required, which I delivered in year two. -'. Five (out of thirty) participants achieved the personal goal they set themselves within the time span they planned. They were interviewed to investigate why they had achieved their goal, and from their feedback the L1FEGIM course contents were updated and the research question was refined to: How does life coaching yourself achieve success? In year three I delivered the revised L1FEGIM course to ten participants from which eight successfully achieved their personal goal. They too, were interviewed. Evaluation of the feedback from the thirteen successful participants was compared, and the common characteristics of their success were identified, and common themes to the study emerged. Three key themes were focused on throughout the research. Firstly, the individual construction of knowledge through the learners' own approach to learning. Secondly life coaching and thirdly success factors. The life coach process used within this research (L1FEGIM) offers a method to goal achievement only if used by committed and persistent individuals who exhibit the specific characteristics of displaying self motivation, a willingness to learn and self belief. The study draws from a variety of sources including the work of Knowles (1973, 1998) where detailed research in andragogy became a credible standard within a'dult learning. The viewpoint adopted in this thesis is that individuals play an active role in the construction of knowledge through their own personalised approach to learning. Thus a constructivist interpretivist framework, is adopted. The results reveal the importance of the individuals' underpinning values and beliefs, and the pull of the socialisation process via L1FEGIM throughout the formal and informal aspects of their learning. Specifically, in using L1FEGIM, a new theoretical view of life coaching emerges from the study suggesting that L1FEGIM, as a holistic process, contributes to realistically identifying and enabling (initially through the support of a Life Coach) self belief towards empowering the individual to persist to achieve their goals.

This dissertation seeks to investigate the term joyful learning as it applies to the learning process. Using methods of narrative inquiry, the collected stories of students serve as the nucleus around which the understanding of joyful learning is built. This inquiry is grounded in the literature of joy in the learning process, whole child learning theory, deep learning, agency in learning, and educational wounds. In this narrative inquiry, 15 students ranging from kindergarten to twelfth grade were asked to tell the story of their experience with learning. A domain analysis was completed on the stories of wounds and celebrations in the interviewees’ varied learning experiences and the identification of commonalities served as a catalyst for an enhancement of education theory and laid the foundation for future research on the importance of joy in the learning process of human beings. The themes emerging in this narrative inquiry were joy, engagement and disengagement in learning, connected learning, motivators and de-motivators, peer influence, positioning, relationship between the student and the teacher, agency, forced learning, standardized assessment and meaningful assessment, educational wounds, and the significance of education. Evidence of educational wounding surfaced in all 15 stories, illuminating a connection between the wounding and the replacement of student-centered learning practices with compulsory, standardized reforms.

The focus of this report is the problem of probabilistic dimensionality reduction and feature learning from high-dimensional data (images). Extracting features and being able to learn from high-dimensional sensory data is an important ability in a general-purpose intelligent system. Dimensionality reduction and feature learning have in the past primarily been done using (convolutional) neural networks or linear mappings, e.g. in principal component analysis. However, these methods do not yield any error bars in the features or predictions. In this report, theory and a model for how dimensionality reduction and feature learning can be done using Gaussian process auto-encoders (GP-AEs) are presented. By using GP-AEs, the variance in the feature space is computed, thus, yielding a measure of the uncertainty in the constructed model. This measure is useful in order to avoid making over-confident system predictions. Results show that GP-AEs are capable of dimensionality reduction and feature learning, but that they suffer from scalability issues and problems with weak gradient signal propagation. Results in reconstruction quality are not as good as those achieved by state-of-the-art methods, and it takes very long to train the model. The model has potential though, since it can scale to large inputs.

This is a practice as research (PAR) study that investigates Process drama as a pedagogical methodology for second language acquisition. The dramas were structured around role play in fictional situations designed to improve the pupils' vocabulary in English, mainly using the strategies of Teacher in Role and Mantle of the Expert. This research demonstrates the effect and affect of placing the pupils in experiential learning environments that empower him/her to be a co-creator of knowledge, as well as the skills needed by the researcher to facilitate this process. The pilot study was designed for implementation over four months, with pupils in second language English classes in grades 2 and 3 (i.e. seven/eight years old). Qualitative research methods were used, including critical and reflexive ethnography, simple questionnaires, and unstructured interviews with teachers and pupils, Teachers' and Visitors' feedback, video recordings and the researcher's journal. The sample of participants was too small to make recommendations, but the study tested the Process Drama method and the techniques of role, Teacher-in-role and Mantle of the Expert. The analysis of the project draws from the theoretical principles applied in other case studies by practitioners in the field, comparing the methods used and their outcomes. As well as the assumption that Process Drama is a good way to teach language because it creates the space for emotion and cognition to co-exist within the learning space.

This study is a case study within Media Technology at Malmö University that explores the gap between design and technology in the Graphic Design course. The study uses the concept of Critical Making and how that can be used to develop a design process learning method in higher education programs, where designing and making artifacts are in focus. The aim is to implement Critical Making in order to allow students to gain deeper understanding and evaluate their material choices and their own reflections in the design process. To achieve the implementation of Critical Making the study uses didactics as a pedagogical approach. The key elements in Critical Making as a design process learning method build on context, reflection, material choice and design process that all form a common design language between students and teachers to exchange knowledge and experiences in future media productions. The common design language is shaped through open assignments, supervision and reflection from the students' material choices and context of the media production they work with during the design process. The findings in the study are also presented as a digitally interactive poster that uses Augmented Reality to show videos and images for the user.

Bibliography: pages 150-158.
Health care systems all over the world are in crisis. The presenting symptom is a cost spiral that is out of control. Money supply is finite, and if this problem continues the system will eventually collapse. There are a number of causes associated with the problem that are usually analysed by reduction, an approach based upon an assumption of simple linear causal relations. This study shows the problem to be the dialectic opposite, in other words these problems are all interrelated through complex causal interactions. Therefore, the health care system is a complex social system and solutions to its problems may be found in terms of the interactions in such a system. An investigation into the history of the health care system shows that the system started with a simple one on one interaction between patients and physicians. At the time of its initiation, very little empirical knowledge was available about illness. After the renaissance, this changed dramatically with a subsequent increase in the ability to diagnose, but also in the complexity to treat illness. However, modern beliefs about illness and illness processes do not reflect the complexity of this knowledge. Beliefs about both illness and knowledge contribute to the process of diagnosis (medical decision making, or problem solving). Furthermore, the expectations, wants, and needs of patients and physicians, as well as the decision environment, increases the complexity and difficulty of this decision making process. These decisions initiate treatment processes that are ultimately represented in the health care system as cost. Therefore, the patient-physician system as the simplest initial interaction is an event that ultimately affects cost. This system is not functioning efficiently at present and a system of inquiry that can improve it may make a contribution to an improved system, and therefore a saving in cost. Altering the diagnostic system from a linear into a circular process, in other words into a learning system, improves both decision making and the use of knowledge. However, an inquiring system is needed in addition that can enhance the rigour of this process. Charles West Churchman devoted a large part of his work to knowledge and the way we acquire knowledge, in other words inquiring systems. His belief is that problem solving ought to be approached in a comprehensive way in order to minimise the risk for making incorrect decisions. Furthermore, because decisions are made upon incomplete information, the solutions will be the cause of new problems. Therefore, problem solving is a never ending cycle of learning. In order to have as complete information as possible about the problem, we have to: know the history of the problem, take a broad view that includes the environment of the problem (use a systems approach), and consider all the alternative solutions to the problem. Virtually all of our knowledge is based upon underlying assumptions. In order to test the validity of the knowledge we use for inquiry and decision making, it is important to test the assumptions upon which the knowledge is based. This is valid in regard to empirical knowledge as well. Finally, according to Churchman, decision making has to be ethical. Therefore, we have to do all we can to ensure that the implementation of the decision will improve the situation, not only now, but also in the future. The application of Churchman's approach to the patient-physician interaction, assists in the synthesis of a more comprehensive world view of health care and illness. This study shows that this leads to important changes in the negative interactions identified as contributing to the health care crisis. In terms of Churchman's approach, the role of physicians can be seen as managers of illness. Their purpose is therefore to plan for the improvement of illness (the problem) in an ethical way. Such planning should include the values of patients in deciding upon appropriate treatment. It is the submission of this study that only a methodology that is able to address complex human systems, such as a systems approach, and a comprehensive philosophy of inquiry, such as that of C West Churchman is appropriate to address the current problems of the health care system.

The study reported here addressed the research question, How do individuals experience the learning and implementing of group process facilitation? It utilised a case-study approach, because such an approach was seen as being particularly useful for exploring the processes and actions that were hard to define for people who were learning and implementing group process facilitation. The study sought to move beyond a reliance on theories of knowledge formulated by positivistic processes, and into the realm of interpretivism and constructivism, in order to explore individual experiences. The literature reviewed indicated that there seemed to be little public knowledge of the implications of learning programs for the individual, from the unique perspective of the individual, particularly as they relate to learning and implementing group process facilitation. It suggested the need for research to address the issue by examining how the participants made sense of, and gave meaning to, their learning experiences. It further suggested the potential value of research into how individuals learning and implementing group process facilitation adapt to their new role, and how they cope with the new realities in their lives as a result of their learning and implementing experiences. The intention in this study was to use a grounded approach that gave adequate recognition to the stories of the participants learning and implementing experiences, and to the meanings attributed by the participants to those experiences. It was intended to draw those stories together so that the commonalities and differences between and among them could be integrated in a reformulated statement of participants' experiences. Thus, a methodology that was both interpretive and constructivist was deemed appropriate, and Denzin's (1989a) 'interpretive interactionism' was selected. As I worked through the analytical phase of the study, I discovered that the chosen methodology did not always provide me with a clear guideline as to how I might follow a logical and coherent process to achieve my intentions. I found that I needed to extend and deepen Denzin's (1989a) model, particularly in the analysis of the data collected. The model (developed and described in Chapter 6) may prove to be useful to other researchers working within a qualitative paradigm. The following substantive findings emerged. Most participants identified the experience of learning and implementing group process facilitation as a turning point in their life and as one that helped them to construct new identities as part of their lived experience. Through reflection and critical reflection, some participants were better able to understand that experience and they were better able to conceptualise it. In light of that conceptualisation, they then made determinations about future work and personal goals. Thus, personal change was an important outcome of their experience. They were more self-aware and more open to learning, although such a process of change involved risk taking. Risk taking sometimes meant that they were fearful or anxious about personal change and were inhibited by self-doubt. Once they worked through their fears, they felt excited about the personal change, and challenged to further their personal learning journey. Participants were also challenged by their experiences of implementing their learning in the workplace. They were able to recognise the significance of the social milieu of the workplace, including the expectations of support about the implementing of their learning from management and peers. That milieu was governed by the novelty of the learning environment, the flexible approach to learning, and the way they, and others, were chosen to attend the learning program. The learning experience helped them to value difference and to recognise the importance of becoming more self-aware, as part of their change process. At times, that process was painful; at other times it was pleasurable - pointing to the paradoxical nature of the learning experience. To undertake the construction of new identities, trust also was recognised as an important issue. Implications for practice are identified within a framework of environmental uncertainty and continuous learning. It is argued that, based on the experiences of the participants, learners may optimise their learning within an appropriate learning environment, and where attention is paid to the use of reflection and critical reflection as part of the learning and development process. Learning programs need to embrace a range of specific activities to enable learners to strengthen their self-belief. Consideration needs to be given to the importance of emotion as part of the learning and development experience. Individuals need to be aware of the importance of the social milieu, and they need to be able to understand their learning experience as an epiphany. For organisations, there is identified a need to attend to learning how to learn. There also appears to be a need for the development of key aspects of organisational functioning to support continual learning. Such aspects include: thinking processes, understanding of the social context in which learning occurs, the need to ensure management and supervisory support for change, and the need to identify and resolve individual, group and organisational conflict. Some participants appeared to be fearful of the reactions of managers and colleagues to their attempts to implement their new learning. Such fears could be overcome at an organisational level by enabling people to apply their learning and to provide meaningful and informative feedback to them. A learning program of the type reported in this study may be useful in helping organisations cope with some of the change processes with which they are faced, and in particular, it may be used to enhance those processes. A number of recommendations for further research may be drawn from this study. Participants in the study reported a transformational change process that has been identified as an epiphany. Further research, therefore, should address the importance of epiphany for the participants in the light of the passage of time. Similar studies need to be conducted in other organisations. The findings of the study are drawn together in two explanatory models: one designated Platforms of Understanding; the other designated Model of Managing Learning in an Organisational Context. Research is needed to explore the interrelationships between the models; the meanings for individuals and organisations of those interrelationships; and how such relationships might be used for the enhancement of learning and development strategies in organisations. Specifically, research focussing on the importance of the historical, political and socio-cultural environment, and into the role played by the organisational and social milieu in which learning and development occurs, needs to be addressed. While there are a number of studies relating to the importance of management and peer support for the transfer of learning in the workplace, research into the cultural significance of such support should be undertaken. Further research should extend to identifying and recognising the impact that such an understanding may have for the culture of the organisation. Learning and implementing group process facilitation is not necessarily a 'behavioural' experience. The part played by emotion in a learning and development program, and later in the organisational context, remains to be addressed. Understanding emotion and emotional responses to organisational change programs, of the type undertaken by participants in the study, should provide significant learning for individuals and organisations as they attempt to cope with organisational change. Most importantly, what this study suggested is that group process facilitation is a social and interactive process that calls for post-positivist research. This study has exposed the richness that may be gained from a constructivist, interpretivist, and progressive-humanist approach in an environment where the identification and recognition of how individuals make sense of, and give meaning to, their learning and implementing experiences, may be seen to have positive organisational and individual benefits. Further research would ideally focus on group process facilitation as the core concept, and be conducted within a constructivist epistemology that reflects progressive-humanistic approaches.
Thesis (PhD Doctorate)
Doctor of Philosophy (PhD)
School of Management
Griffith Business School
Full Text

Gaussian processes are powerful nonparametric distributions over continuous functions that have become a standard tool in modern probabilistic machine learning. However, the applicability of Gaussian processes in the large-data regime and in hierarchical probabilistic models is severely limited by analytic and computational intractabilities. It is, therefore, important to develop practical approximate inference and learning algorithms that can address these challenges. To this end, this dissertation provides a comprehensive and unifying perspective of pseudo-point based deterministic approximate Bayesian learning for a wide variety of Gaussian process models, which connects previously disparate literature, greatly extends them and allows new state-of-the-art approximations to emerge. We start by building a posterior approximation framework based on Power-Expectation Propagation for Gaussian process regression and classification. This framework relies on a structured approximate Gaussian process posterior based on a small number of pseudo-points, which is judiciously chosen to summarise the actual data and enable tractable and efficient inference and hyperparameter learning. Many existing sparse approximations are recovered as special cases of this framework, and can now be understood as performing approximate posterior inference using a common approximate posterior. Critically, extensive empirical evidence suggests that new approximation methods arisen from this unifying perspective outperform existing approaches in many real-world regression and classification tasks. We explore the extensions of this framework to Gaussian process state space models, Gaussian process latent variable models and deep Gaussian processes, which also unify many recently developed approximation schemes for these models. Several mean-field and structured approximate posterior families for the hidden variables in these models are studied. We also discuss several methods for approximate uncertainty propagation in recurrent and deep architectures based on Gaussian projection, linearisation, and simple Monte Carlo. The benefit of the unified inference and learning frameworks for these models are illustrated in a variety of real-world state-space modelling and regression tasks.

Machine learning and process mining are two techniques that are becoming more and more popular among organisations for business intelligence purposes. Results from these techniques can be very useful for organisations' decision-making. The Swedish National Forensic Centre (NFC), an organisation that performs forensic analyses, is in need of a way to visualise and understand its administration process. In addition, the organisation would like to be able to predict the time analyses will take to perform. In this project, it was evaluated if machine learning and process mining could be used on NFC's administration process-related data to satisfy the organisation's needs. Using the process mining tool Mehrwerk Process Mining implemented in the software Qlik Sense, different process variants were discovered from the data and visualised in a comprehensible way. The process variants were easy to interpret and useful for NFC. Machine learning regression models were trained on the data to predict analysis length. Two different datasets were tried, a large dataset with few features and a smaller dataset with more features. The models were then evaluated on test datasets. The models did not predict the length of analyses in an acceptable way. A reason to this could be that the information in the data was not sufficient for this prediction.

In the recent years, the growing number of recorded events made the interest in process mining techniques expand. These techniques make it possible to learn the model of a process, to compare a recent event log with an existing model or to enhance the process model using the information extracted from the log. Most of the existing process mining algorithms only make use of positive examples of a business process in order to extract its model, however, negative ones can bring major benefits. In this work, a discovery algorithm, inspired by the one presented by Mooney in 1995, that takes advantage of both positive and negative sequences of actions is presented in two different versions that return a declarative model connected respectively in disjunctive and conjunctive logic formulas.

Thesis (Ph.D.)--Antioch University, 2007.
Title from PDF t.p. (viewed Apr. 29, 2008). Advisor: Carolyn Kenny, PhD. "Submitted to the Ph.D. in Leadership & Change Program of Antioch University in partial fulfillment of the requirements for the degree of Doctor of Philosophy October, 2007"--from the title page. Keywords: transformative learning, middle-age, portraiture; artists, phenomenology, Jungian psychology, midlife, depth psychology, life change. Includes bibliographical references (p. 265-273).

CONSELHO NACIONAL DE DESENVOLVIMENTO CIENTÍFICO E TECNOLÓGICO
FUNDAÇÃO DE APOIO À PESQUISA DO ESTADO DO RIO DE JANEIRO
Neste trabalho foram desenvolvidos e avaliados métodos com o objetivo de melhorar e acelerar o processo de aprendizado do modelo de Reinforcement Learning Neuro-Fuzzy Hierárquico Politree (RL-NFHP). Este modelo pode ser utilizado para dotar um agente de inteligência através de processo de Aprendizado por Reforço (Reinforcement Learning). O modelo RL-NFHP apresenta as seguintes características: aprendizado automático da estrutura do modelo; auto-ajuste dos parâmetros associados à estrutura; capacidade de aprendizado da ação a ser adotada quando o agente está em um determinado estado do ambiente; possibilidade de lidar com um número maior de entradas do que os sistemas neuro-fuzzy tradicionais; e geração de regras linguísticas com hierarquia. Com intenção de melhorar e acelerar o processo de aprendizado do modelo foram implementadas seis políticas de seleção, sendo uma delas uma inovação deste trabalho (Q-DC-roulette); implementado o método early stopping para determinação automática do fim do treinamento; desenvolvido o eligibility trace cumulativo; criado um método de poda da estrutura, para eliminação de células desnecessárias; além da reescrita do código computacional original. O modelo RL-NFHP modificado foi avaliado em três aplicações: o benchmark Carro na Montanha simulado, conhecido na área de agentes autônomos; uma simulação robótica baseada no robô Khepera; e uma num robô real NXT. Os testes efetuados demonstram que este modelo modificado se ajustou bem a problemas de sistemas de controle e robótica, apresentando boa generalização. Comparado o modelo RL-NFHP modificado com o original, houve aceleração do aprendizado e obtenção de menores modelos treinados.
In this work, methods were developed and evaluated in order to improve and accelerate the learning process of Reinforcement Learning Neuro-Fuzzy Hierarchical Politree Model (RL-NFHP). This model is employed to provide an agent with intelligence, making it autonomous, due to the capacity of ratiocinate (infer actions) and learning, acquired knowledge through interaction with the environment by Reinforcement Learning process. The RL-NFHP model has the following features: automatic learning of structure of the model; self-adjustment of parameters associated with its structure, ability to learn the action to be taken when the agent is in a particular state of the environment; ability to handle a larger number of inputs than the traditional neuro-fuzzy systems; and generation of rules with linguistic interpretable hierarchy. With the aim to improve and accelerate the learning process of the model, six selection action policies were developed, one of them an innovation of this work (Q-DC-roulette); implemented the early stopping method for automatically determining the end of the training; developed a cumulative eligibility trace; created a method of pruning the structure, for removing unnecessary cells; in addition to rewriting the original computer code. The modified RL-NFHP model was evaluated in three applications: the simulated benchmark Car-Mountain problem, well known in the area of autonomous agents; a simulated application in robotics based on the Khepera robot; and an application in a real robot. The experiments show that this modified model fits well the problems of control systems and robotics, with a good generalization. Compared the modified RL-NFHP model with the original one, there was acceleration of learning process and smaller structures of the model trained.