I dagens grundskola ställs en mängd olika krav på elever och på de kunskaper som bör besittas. Det beror på att skolan har till uppgift att förbereda eleverna inför det föränderliga samhälle vi lever i. För att kunna göra detta krävs att de är aktiva och delaktiga i den undervisning som rör dem och för att öka denna delaktighet finns en mängd olika arbetsmodeller som alla handlar om att förbättra elevers aktivitet. Bland dessa finns en modell som kallas problembaserat lärande och förkortas PBL. Syftet med det här examensarbetet är att undersöka om, och i så fall hur ett problembaserat lärande kan användas i grundskolans tidigare år som en elevaktiv undervisningsmodell. Anledningen till att detta ämne och syfte valts är för att merparten av den forskning som finns som rör PBL fokuserar på högre studier medan de lägre undervisningsnivåerna har undersökts relativt lite. I ett försök att minska kunskapsglappet har denna undersökning syftat till att påvisa hur lärare arbetar och tidigare har arbetat med problembaserat lärande i grundskolans tidigare år. Denna undersökning har genomförts med hjälp av kvalitativt inriktade interjuver vilka resulterade i slutsatserna att det är möjligt att arbeta med PBL som en elevaktiv undervisningsmodell i grundskolan och dess tidigare år samt att eleverna tycks bli mer aktiva och delaktiga. Det är grundat på resultatet av undersökningen där några av personerna som deltog ansåg sig arbeta eller tidigare ha arbetat med problembaserat lärande och gjort detta på ett tillfredsställande sätt. Dessutom tycktes eleverna, enligt undersökningspersonerna, bli mer aktiva och delaktiga än i traditionell undervisning.

In this dissertation, the problem of learning from highly imbalanced data is studied. Imbalance data learning is of great importance and challenge in many real applications. Dealing with a minority class normally needs new concepts, observations and solutions in order to fully understand the underlying complicated models. We try to systematically review and solve this special learning task in this dissertation.We propose a new ensemble learning framework—Diversified Ensemble Classifiers for Imbal-anced Data Learning (DECIDL), based on the advantages of existing ensemble imbalanced learning strategies. Our framework combines three learning techniques: a) ensemble learning, b) artificial example generation, and c) diversity construction by reversely data re-labeling. As a meta-learner, DECIDL utilizes general supervised learning algorithms as base learners to build an ensemble committee. We create a standard benchmark data pool, which contains 30 highly skewed sets with diverse characteristics from different domains, in order to facilitate future research on imbalance data learning. We use this benchmark pool to evaluate and compare our DECIDL framework with several ensemble learning methods, namely under-bagging, over-bagging, SMOTE-bagging, and AdaBoost. Extensive experiments suggest that our DECIDL framework is comparable with other methods. The data sets, experiments and results provide a valuable knowledge base for future research on imbalance learning. We develop a simple but effective artificial example generation method for data balancing. Two new methods DBEG-ensemble and DECIDL-DBEG are then designed to improve the power of imbalance learning. Experiments show that these two methods are comparable to the state-of-the-art methods, e.g., GSVM-RU and SMOTE-bagging. Furthermore, we investigate learning on imbalanced data from a new angle—active learning. By combining active learning with the DECIDL framework, we show that the newly designed Active-DECIDL method is very effective for imbalance learning, suggesting the DECIDL framework is very robust and flexible.Lastly, we apply the proposed learning methods to a real-world bioinformatics problem—protein methylation prediction. Extensive computational results show that the DECIDL method does perform very well for the imbalanced data mining task. Importantly, the experimental results have confirmed our new contributions on this particular data learning problem.

This study aimed to investigate the potentials of two cognitive tools for instructional planning (Instructional Planning Self-reflective Tool, IPSRT, and Constructivist Planning Self-reflective Tool, CPSRT) and microteaching in gaining instructional planning and teaching skills for preservice teachers. The participants were 51 fourth year students in Computer Education and Instructional Technology program. The study is an action research with three main foci. The first focus of this study aimed at investigating contributions and challenges involved in the use of the cognitive tools for instructional planning with tutoring from the instructor. More specifically, to what extent the preservice teachers followed these tools during this process, the effects of these tools on preservice teachers&rsquo / self-efficacy, the perceived instrumentality regarding instructional planning, and the perceived contributions and challenges presented by these tools were focused. Both tools were introduced to the two sections, in different orders within four weeks. The data for this focus were collected by means of questionnaires, interviews and documents (lesson plans). This focus revealed that / expect for writing objectives, the participants could make instructional plans according to the IPSRT. They could also follow the CPSRT to design the instructional goal, required characteristics of learning activities and the assessment. Both tools were found to significantly increase their initial self-efficacy beliefs. They found CPSRT more flexible, while IPSRT easier and more helpful. This focus indicated that IPSRT and CPSRT can be used as supportive tools in preservice teachers&rsquo / gaining instructional planning skills. If both tools were used, it would be better to introduce IPSRT at first and then CPSRT. The second focus of this study was to explore the contributions and challenges of microteaching activities regarding preservice teachers&rsquo / instructional planning and teaching skills. The microteaching activities took eight weeks. Throughout this phase, each student planned a 20-minute microteaching with tutoring from the instructor and performed it in the classroom. The performers were formatively evaluated through a microteaching assessment form by the instructor, the teaching assistants and some preservice teachers. Then the performers made a self-reflection assignment about their microteaching performance, considering those evaluations. In the following semester, 15 participants&rsquo / perceptions about the contributions and challenges posed by microteaching activities for their instructional planning and teaching skills were obtained through interviews. More specifically, their perceptions about the microteaching planning process with tutoring, performing microteaching, formatively assessing peers&rsquo / microteaching performances, being assessed by peers, and doing self-reflection assignment were analyzed. This focus revealed that although preservice teachers perceive microteaching activities as valuable experiences, microteaching would be more beneficial if the pupils were real ones, not their class-mates. The third focus was to investigate the effects of the cognitive tools and microteaching activities on preservice teachers&rsquo / lesson planning and teaching skills in their field teaching. For this aim, 12 participants&rsquo / field teaching lesson plans and their performance assessments were analyzed. It was found that many of them preferred using the Microteaching Planning Guide and they had no difficulty in their lesson planning. As to field teaching performance, the analyses of the assessment forms showed that a majority of them performed successfully. Besides, most of them were observed not to have anxiety during field teaching. This focus showed that these cognitive tools and microteaching activities could improve preservice teachers&rsquo / self-confidence in lesson planning and teaching skills in real class environment. Considering to meeting the need for better qualified teachers, this study promised that applying these cognitive tools and microteaching model in schools of teacher education is likely to contribute to the instructional planning and teaching skills of preservice computer teachers. This study also offers suggestive implications for how to improve teaching methods courses with the two cognitive tools and microteaching, as well.

Not available / text

published_or_final_version / Education / Master / Master of Education

Experiencing Freefall is an inquiry into outdoor experiencing. It focuses on both my experiences with a group of Grade 10 students in an outdoor adventure education program and my personal experiencing of the outdoors. I explore the awareness we embody within moments of unexpected happenings while negotiating Whitewater rapids or searching for a handhold while clinging to the side of a cliff face. Also in this thesis I explore the 'phenomena of experiencing' which emerges out of our actions and interactions within outdoor activities. The challenge of this dissertation is capturing in prose, the phenomena of experiencing and 'embodied awareness' arising through such unexpected instances. Hence, the thesis, in trying to articulate the complexity of experiencing in the outdoors, uses stories, poetry and the metaphor of life, breath, and mountaineering to invite the reader on a journey of inquiry. This thesis escorts the reader, like a true pedagogue, into an outdoor environment of experiencing that opens the reader to ponder pedagogical possibilities. I explore several themes in the thesis which include 'freefall,' community, 'turning points,' and 'embodied respect' using a methodology of 'enactive inquiry.' The thesis takes a journey through each theme by weaving students' stories from the study, my own personal stories of the unexpected, and the theory of enaction. The thesis creates an opportunity for readers to embrace their own struggles, fears, and inquiry. Through the use of outdoor stories to illustrate moments of freefall into the unfamiliar or unknown, we may imagine pedagogical possibilities. As an enactive inquiry, this research thesis embodies an "education" or way of being, living, experiencing that explores unexpected happenings. In articulating an ecological perspective of experiencing, the thesis juxtaposes encounters in the outdoors with enactive theory to move beyond traditional representationalist models of cognition. Specifically, I focus on the embodied awareness that arises through phenomena of experiencing and its relation to pedagogy. The thesis contributes to the theory of the enactive approach by bringing examples of human experience which unfold, not only our interactions within the ecological web of the outdoor world, but an emergent space of pedagogical possibilities. As such, this thesis is an experiential work through which the reader may realize their own interpreting of possible pedagogies for many educational contexts.

This is a case study of people living in the Merebank/Wentworth area which is highly polluted. This area is sandwiched between the engine refinery SAPREF, Mondi paper mill, the Durban airport and other small industries. As such it is exposed to a mixture of gases in the atmosphere which is detrimental to the health of the residents. The residents have discovered that they cannot rely on government and authorities to bring relief to the situation and have thus made it a point to acquire 'really useful knowledge' in making industries accept accountability and "clean up their act." The study has confirmed that community organisations played a significant role in raising awareness of the pollution issue and mobilising people in social action which has resulted in the industries being pressurised to improve technology in refining crude oil. The study aimed to explore the kinds of learning that took place when the residents collectively participated in social action. Social interaction is a salient feature of learning. The case study is 'heuristic' in nature in that the community gives new meaning to their experience. A situated learning approach based on social learning theory is proposed as a theoretical framework for the study. Data for the inquiry into the participants group learning and social action consisted of taped interviews, participant observation and analysis of documents. / Thesis (M.Ed.)-University of Natal, Durban, 2002.

The 1996 constitution of South Africa was adopted as the supreme law of the Republic so as to establish a new society based on democratic values, to 'improve the lives of all citizens and to free the potential of all persons by every means possible' (1996:Section 27). Every person now has certain inherent rights which were denied to most prior to the 1994 elections. All persons have the right to dignity, and the right to have their dignity respected and protected. The State agrees, 'within its resources as outlined in its macro economic strategy GEAR' (Beck 2000: 195) to take reasonable legislative and other measures to achieve the progressive realization of people's rights and to have these rights respected. There is a major shift in the way society is governed. Government legislation reflects the move away from the harsh, discriminatory laws of the past, to a new social order based on democratic principles. Most welfare organizations are willing to embrace the new dispensation and some are well advanced in the transformation process which embraces the developmental approach to social welfare. This research looks at two such organizations within the context of a case study. Its purpose is not to detail the difficulties and tensions faced by the organizations in terms of the implementation of a developmental approach to social welfare, but rather to explore how two groups of people from very diverse backgrounds, politically, historically and economically, learn to work together on a developmental project during a time of monumental change. It details how the two organisations made progress together in spite of their many difficulties and differences, to bring each phase of the Project to fruition during the period October 1997 - October 2001. I use the actual geographical names of the Project during the research but the names of the organisations and the participants have been changed to protect identities. / Thesis (M.Ed.)-University of KwaZulu-Natal, Durban, 2003.

Active Learning arises as an important issue in several supervised learning scenarios where obtaining data is cheap, but labeling is costly. In general, this consists in a query strategy, a greedy heuristic based on some selection criterion, which searches for the potentially most informative observations to be labeled in order to form a training set. A query strategy is therefore a biased sampling procedure since it systematically favors some observations by generating biased training sets, instead of making independent and identically distributed draws. The main hypothesis of this thesis lies in the reduction of the bias inherited from the selection criterion. The general proposal consists in reducing the bias by selecting the minimal training set from which the estimated probability distribution is as close as possible to the underlying distribution of overall observations. For that, a novel general active learning query strategy has been developed using an Information-Theoretic framework. Several experiments have been performed in order to evaluate the performance of the proposed strategy. The obtained results confirm the hypothesis about the bias, showing that the proposal outperforms the baselines in different datasets.

The field of Machine Learning is concerned with the development of algorithms, models and techniques that solve challenging computational problems by learning from data representative of the problem (e.g. given a set of medical images previously classified by a human expert, build a model to predict unseen images as either benign or malignant). Many important real-world problems have been formulated as supervised learning problems. The assumption is that a data set is available containing the correct output (e.g. class label or target value) for each given data point. In many application domains, obtaining the correct outputs (labels) for data points is a costly and time-consuming task. This has provided the motivation for the development of Machine Learning techniques that attempt to minimize the number of labeled data points while maintaining good generalization performance on a given problem. Active Learning is one such class of techniques and is the focus of this thesis. Active Learning algorithms select or generate unlabeled data points to be labeled and use these points for learning. If successful, an Active Learning algorithm should be able to produce learning performance (e.g test set error) comparable to an equivalent supervised learner using fewer labeled data points. Theoretical, algorithmic and experimental Active Learning research has been conducted and a number of successful applications have been demonstrated. However, the scope of many of the experimental studies on Active Learning has been relatively small and there are very few large-scale experimental evaluations of Active Learning techniques. A significant amount of performance variability exists across Active Learning experimental results in the literature. Furthermore, the implementation details and effects of experimental factors have not been closely examined in empirical Active Learning research, creating some doubt over the strength and generality of conclusions that can be drawn from such results. The Active Learning model/system used in this thesis is the Optimal Active Learning algorithm framework with Gaussian Processes for regression problems (however, most of the research questions are of general interest in many other Active Learning scenarios). Experimental and implementation details of the Active Learning system used are described in detail, using a number of regression problems and datasets of different types. It is shown that the experimental results of the system are subject to significant variability across problem datasets. The hypothesis that experimental factors can account for this variability is then investigated. The results show the impact of sampling and sizes of the datasets used when generating experimental results. Furthermore, preliminary experimental results expose performance variability across various real-world regression problems. The results suggest that these experimental factors can (to a large extent) account for the variability observed in experimental results. A novel resampling technique for Optimal Active Learning, called '3-Sets Cross-Validation', is proposed as a practical solution to reduce experimental performance variability. Further results confirm the usefulness of the technique. The thesis then proposes an extension to the Optimal Active Learning framework, to perform learning via membership queries via a novel algorithm named MQOAL. The MQOAL algorithm employs the Metropolis-Hastings Markov chain Monte Carlo (MCMC) method to sample data points for query selection. Experimental results show that MQOAL provides comparable performance to the pool-based OAL learner, using a very generic, simple MCMC technique, and is robust to experimental factors related to the MCMC implementation. The possibility of making queries in batches is also explored experimentally, with results showing that while some performance degradation does occur, it is minimal for learning in small batch sizes, which is likely to be valuable in some real-world problem domains.