Rozprawy doktorskie na temat „Learning with pre-built models”

Doctor of Philosophy (PhD)
Environmental issues are complex and understanding them involves integration of different areas of knowledge, feedback and time delays, however strategies to cope with complexity are not often used or taught in environmental education. The aim of this thesis is to examine the benefit of three such strategies for environmental education: multiple external representations, learning from models, and collaborative learning. The socio-environmental system modelled was visitor impact in a national park in Australia. Students in Year 9 and 10 from two schools were given a text description (Text group) and either a system dynamics model (SDM group), an agent-based model (ABM group), or both models (SDM & ABM group). This experimental design allowed learning outcomes (environmental and system dynamics knowledge, and understanding of the socio-environmental system) and use of the model(s) (in terms of the proportion of time spent on each screen, activities, and strategies) to be compared in each learning environment (individual and collaborative). Multiple external representations were the most successful strategy in the individual learning environment in terms of increases in environmental knowledge. However, students given only the system dynamics model had greater understanding of the system, and students given only the agent-based model increased environmental knowledge easily identified in the animated representation. Prior knowledge, patterns of use, strategies for changing variables and the representational affordances of the models explained some of these differences. In particular, prior knowledge was an important indicator of how students coordinated use of the models in the SDM & ABM group. Learning with a system dynamics model was the most successful strategy for students in the collaborative learning environment. Differences between the learning environments were detected in all groups with respect to both learning outcomes and use of the models due to prior knowledge, interrogation of the models, and the learning environments themselves. These experiments have provided evidence that strategies for understanding complex systems provide viable methods of communicating complex ideas to school-aged students with varying levels of prior knowledge. In particular, multiple external representations provided students with flexibility in how they learned; models allowed students to experiment with a system otherwise not allowed; and a collaborative learning environment facilitated students’ interpretation of a system dynamics model.

Environmental issues are complex and understanding them involves integration of different areas of knowledge, feedback and time delays, however strategies to cope with complexity are not often used or taught in environmental education. The aim of this thesis is to examine the benefit of three such strategies for environmental education: multiple external representations, learning from models, and collaborative learning. The socio-environmental system modelled was visitor impact in a national park in Australia. Students in Year 9 and 10 from two schools were given a text description (Text group) and either a system dynamics model (SDM group), an agent-based model (ABM group), or both models (SDM & ABM group). This experimental design allowed learning outcomes (environmental and system dynamics knowledge, and understanding of the socio-environmental system) and use of the model(s) (in terms of the proportion of time spent on each screen, activities, and strategies) to be compared in each learning environment (individual and collaborative). Multiple external representations were the most successful strategy in the individual learning environment in terms of increases in environmental knowledge. However, students given only the system dynamics model had greater understanding of the system, and students given only the agent-based model increased environmental knowledge easily identified in the animated representation. Prior knowledge, patterns of use, strategies for changing variables and the representational affordances of the models explained some of these differences. In particular, prior knowledge was an important indicator of how students coordinated use of the models in the SDM & ABM group. Learning with a system dynamics model was the most successful strategy for students in the collaborative learning environment. Differences between the learning environments were detected in all groups with respect to both learning outcomes and use of the models due to prior knowledge, interrogation of the models, and the learning environments themselves. These experiments have provided evidence that strategies for understanding complex systems provide viable methods of communicating complex ideas to school-aged students with varying levels of prior knowledge. In particular, multiple external representations provided students with flexibility in how they learned; models allowed students to experiment with a system otherwise not allowed; and a collaborative learning environment facilitated students’ interpretation of a system dynamics model.

After decades of research, a host of analytical difficulties is still hindering our understanding of the influences of the built form on travel. The main challenges are (a) assembling good quality data that reflects the majority of the known influences and that supports continuous monitoring, and (b) making sense methodologically of the many variables which strongly intercorrelate. This study uses the UK national travel survey (NTS) data that is among the most comprehensive of its form in the world. The fact that it has rarely been used so far for this purpose may be attributable to the methodological difficulties. This dissertation aims to develop a new analytical framework based on extended structural equation models (SEMs) in order to overcome some of the key methodological difficulties in quantifying the influences of the built form on travel, and in addition to provide a means to continuously monitor any changes in the effects over time. The analyses are focused on employed adults, because they are not only the biggest UK population segment with the highest per capita travel demand, but also the segment that are capable of adapting more rapidly to changing land use, built form and transport supply conditions. The research is pursued through three new models. Model 1 is a path diagram coupled with factor analyses, which estimates continuous, categorical and binary dependent variables. The model estimates the influences on travel distance, time and trip frequency by trip purpose while accounting for self-selection, spatial sorting, endogeneity of car ownership, and interactions among trip purposes. The results highlight stark differences among commuters, particularly the mobility disadvantages of women, part time and non-car owning workers even when they live in the most accessible urban areas. Model 2 incorporates latent categorisation analyses in order to identify a tangible typology of the built form and the associated variations in impacts on travel. Identifying NTS variables as descriptors for tangible built form categories provides an improved basis for investigating land use and transport planning interventions. The model reveals three distinct built form categories in the UK with striking variations in the patterns of influences. Model 3 further investigates the variations across the built form categories. The resulting random intercept SEM provides a more precise quantification of the influences of self-selection and spatial sorting across the built form categories for each socioeconomic group. Four research areas are highlighted for further studies: First, new preference, attitude and behavioural parameters may be introduced through incorporating non-NTS behavioural surveys; Second, the new SEMs provide a basis for incorporating choice modelling where the utility function is defined with direct, indirect and latent variables; Third, conceptual and methodological developments – such as non-parametric latent class analysis, allow expanding the current model to monitor changes in travel behaviour as and when new NTS or non NTS data become available. Fourth, the robustness of the inferences regarding causal or directional influences may require further quantification through designing new panel data sets, building on the findings above.

Metal organic frameworks (MOFs) are a class of nanoporous materials composed through self-assembly of inorganic and organic structural building units (SBUs). MOFs show great promise for many applications due to their record-breaking internal surface areas and tunable pore chemistry. This thesis work focuses on gas separation applications of MOFs in the context of carbon capture and storage (CCS) technologies. CCS technologies are expected to play a key role in the mitigation of anthropogenic CO2 emissions in the near future. In the first part of the thesis, robust machine learning quantitative structure-property relationship (QSPR) models are developed to predict CO2 working capacity and CO2/H2 selectivity for pre-combustion carbon capture using the most topologically diverse database of hypothetical MOF structures constructed to date (358,400 MOFs, 1166 network topologies). The support vector regression (SVR) models are developed on a training set of 35,840 MOFs (10% of the database) and validated on the remaining 322,560 MOFs. The most accurate models for CO2 working capacities (R2 = 0.944) and CO2/H2 selectivities (R2 = 0.876) are built from a combination of six geometric descriptors and three novel y-range normalized atomic-property-weighted radial distribution function (AP-RDF) descriptors. 309 common MOFs are identified between the grand canonical Monte Carlo (GCMC) calculated and SVR-predicted top-1000 high-performing MOFs ranked according to a normalized adsorbent performance score. This work shows that SVR models can indeed account for the topological diversity exhibited by MOFs. In the second project of this thesis, computational simulations are performed on a MOF, CALF-20, to examine its chemical and physical properties which are linked to its exceptional water-resisting ability. We predict the atomic positions in the crystal structure of the bulk phase of CALF-20, for which only a powder X-ray diffraction pattern is available, from a single crystal X-ray diffraction pattern of a metastable phase of CALF-20. Using the predicted CALF-20 structure, we simulate adsorption isotherms of CO2 and N2 under dry and humid conditions which are in excellent agreement with experiment. Snapshots of the CALF-20 undergoing water sorption simulations reveal that water molecules in a given pore adsorb and desorb together due to hydrogen bonding. Binding sites and binding energies of CO2 and water in CALF-20 show that the preferential CO2 uptake at low relative humidities is driven by the stronger binding energy of CO2 in the MOF, and the sharp increase in water uptake at higher relative humidities is driven by the strong intermolecular interactions between water. In the third project of this thesis, we use computational simulations to investigate the effects of residual solvent on Ni-BPM’s CH4 and N2 adsorption properties. Single crystal X-ray diffraction data shows that there are two sets of positions (Set 1 and 2) that can be occupied by the 10 residual DMSO molecules in the Ni-BPM framework. GCMC simulations of CH4 and N2 uptake in Ni-BPM reveal that CH4 uptake is in closest agreement with experiment when the 10 DMSO’s are placed among the two sets of positions in equal ratio (Mixed Set). Severe under-prediction and over-prediction of CH4 uptake are observed when the DMSO’s are placed in Set1 and Set 2 positions, respectively. Through binding site analysis, the CH4 binding sites within the Ni-BPM framework are found to overlap with the Set 1 DMSO positions but not with the Set 2 DMSO positions which explains the deviations in CH4 uptake observed for these cases. Binding energy calculations reveal that CH4 molecules are most stabilized when the DMSO’s are in the Mixed Set of positions.

The Natural Language Processing (NLP) research area has seen notable advancements in recent years, one being the ELECTRA model which improves the sample efficiency of BERT pre-training by introducing a discriminative pre-training approach. Most publicly available language models are trained on general-domain datasets. Thus, research is lacking for niche domains with domain-specific vocabulary. In this paper, the process of adapting a BERT-like model to the telecom domain is investigated. For efficiency in training the model, the ELECTRA approach is selected. For measuring target- domain performance, the Question Answering (QA) downstream task within the telecom domain is used. Three domain adaption approaches are considered: (1) continued pre- training on telecom-domain text starting from a general-domain checkpoint, (2) pre-training on telecom-domain text from scratch, and (3) pre-training from scratch on a combination of general-domain and telecom-domain text. Findings indicate that approach 1 is both inexpensive and effective, as target- domain performance increases are seen already after small amounts of training, while generalizability is retained. Approach 2 shows the highest performance on the target-domain QA task by a wide margin, albeit at the expense of generalizability. Approach 3 combines the benefits of the former two by achieving good performance on QA both in the general domain and the telecom domain. At the same time, it allows for a tokenization vocabulary well-suited for both domains. In conclusion, the suitability of a given domain adaption approach is shown to depend on the available data and computational budget. Results highlight the clear benefits of domain adaption, even when the QA task is learned through behavioral fine-tuning on a general-domain QA dataset due to insufficient amounts of labeled target-domain data being available.
Dubbelriktade språkmodeller som BERT har på senare år nått stora framgångar inom språkteknologiområdet. Flertalet vidareutvecklingar av BERT har tagits fram, bland andra ELECTRA, vars nyskapande diskriminativa träningsprocess förkortar träningstiden. Majoriteten av forskningen inom området utförs på data från den allmänna domänen. Med andra ord finns det utrymme för kunskapsbildning inom domäner med områdesspecifikt språk. I detta arbete utforskas metoder för att anpassa en dubbelriktad språkmodell till telekomdomänen. För att säkerställa hög effektivitet i förträningsstadiet används ELECTRA-modellen. Uppnådd prestanda i måldomänen mäts med hjälp av ett frågebesvaringsdataset för telekom-området. Tre metoder för domänanpassning undersöks: (1) fortsatt förträning på text från telekom-området av en modell förtränad på den allmänna domänen; (2) förträning från grunden på telekom-text; samt (3) förträning från grunden på en kombination av text från telekom-området och den allmänna domänen. Experimenten visar att metod 1 är både kostnadseffektiv och fördelaktig ur ett prestanda-perspektiv. Redan efter kort fortsatt förträning kan tydliga förbättringar inom frågebesvaring inom måldomänen urskiljas, samtidigt som generaliserbarhet kvarhålls. Tillvägagångssätt 2 uppvisar högst prestanda inom måldomänen, om än med markant sämre förmåga att generalisera. Metod 3 kombinerar fördelarna från de tidigare två metoderna genom hög prestanda dels inom måldomänen, dels inom den allmänna domänen. Samtidigt tillåter metoden användandet av ett tokenizer-vokabulär väl anpassat för båda domäner. Sammanfattningsvis bestäms en domänanpassningsmetods lämplighet av den respektive situationen och datan som tillhandahålls, samt de tillgängliga beräkningsresurserna. Resultaten påvisar de tydliga vinningar som domänanpassning kan ge upphov till, även då frågebesvaringsuppgiften lärs genom träning på ett dataset hämtat ur den allmänna domänen på grund av otillräckliga mängder frågebesvaringsdata inom måldomänen.

As the information flow on the Internet keeps growing it becomes increasingly easy to miss important news which does not have a mass appeal. Combating this problem calls for increasingly sophisticated information retrieval methods. Pre-trained transformer based language models have shown great generalization performance on many natural language processing tasks. This work investigates how well such a language model, Open AI’s General Pre-trained Transformer 2 model (GPT-2), generalizes to information retrieval and classification of online news articles, written in English, with the purpose of comparing this approach with the more traditional method of Term Frequency-Inverse Document Frequency (TF-IDF) vectorization. The aim is to shed light on how useful state-of-the-art transformer based language models are for the construction of personalized information retrieval systems. Using transfer learning the smallest version of GPT-2 is trained to rank and classify news articles achieving similar results to the purely TF-IDF based approach. While the average Normalized Discounted Cumulative Gain (NDCG) achieved by the GPT-2 based model was about 0.74 percentage points higher the sample size was too small to give these results high statistical certainty.
Informationsflödet på Internet fortsätter att öka vilket gör det allt lättare att missa viktiga nyheter som inte intresserar en stor mängd människor. För att bekämpa detta problem behövs allt mer sofistikerade informationssökningsmetoder. Förtränade transformermodeller har sedan ett par år tillbaka tagit över som de mest framstående neurala nätverken för att hantera text. Det här arbetet undersöker hur väl en sådan språkmodell, Open AIs General Pre-trained Transformer 2 (GPT-2), kan generalisera från att generera text till att användas för informationssökning och klassificering av texter. För att utvärdera detta jämförs en transformerbaserad modell med en mer traditionell Term Frequency- Inverse Document Frequency (TF-IDF) vektoriseringsmodell. Målet är att klargöra hur användbara förtränade transformermodeller faktiskt är i skapandet av specialiserade informationssökningssystem. Den minsta versionen av språkmodellen GPT-2 anpassas och tränas om till att ranka och klassificera nyhetsartiklar, skrivna på engelska, och uppnår liknande prestanda som den TF-IDF baserade modellen. Den GPT-2 baserade modellen hade i genomsnitt 0.74 procentenheter högre Normalized Discounted Cumulative Gain (NDCG) men provstorleken var ej stor nog för att ge dessa resultat hög statistisk säkerhet.

Streams in small watersheds are often known to exhibit diel fluctuations, in which streamflow oscillates on a 24-hour cycle. Streamflow diel fluctuations, which we investigate in this study, are an informative indicator of environmental processes. However, in Environmental Data sets, as well as many others, there is a range of noise associated with individual data points. Some points are extracted under relatively clear and defined conditions, while others may include a range of known or unknown confounding factors, which may decrease those points' validity. These points may or may not remain useful for training, depending on how much uncertainty they contain. We submit that in situations where some variability exists in the clarity or 'Confidence' associated with individual data points – Notably environmental data – an approach that factors this confidence into account during the training phase is beneficial. We propose a methodological framework for assigning confidence to individual data records and augmenting training with that information. We then exercise this methodology on two separate datasets: A simulated data set, and a real-world, Environmental Science data set with a focus on streamflow diel signals. The simulated data set provides integral understanding of the nature of the data involved, and the Environmental Science data set provides a real-world case study of an application of this methodology against noisy data. Both studies' results indicate that applying and utilizing confidence in training increases performance and assists in the Data Mining Process.

Flipped learning has received increased recognition as an innovative pedagogical approach that has the potential to improve students’ learning experience in higher education. This approach creates a ‘reversed’ learning experience, where portions of the didactic lecture traditionally presented in class is moved online in the form of pre-learning materials. There is increasing evidence that this leads to improvements in academic performance with the online pre-learning materials being an underlying factor. This thesis reports student behavioural engagement, behavioural patterns, and approaches to learning with the online component of a partially flipped learning model and its impact on student academic performance in chemistry courses. An engagement index was developed to quantify student engagement levels with pre-learning materials. The findings revealed higher levels of engagement led to significant improvements in academic performance. Several patterns were detected when measuring students’ frequency of access for each of the pre-learning materials. The dominant pattern revealed that students tend to favour accessing a pre-learning quiz more frequently than the video. Most students self-identified to be strategic learners and were categorised to be moderately or highly engaged with a preference to accessing the quizzes more frequently than the videos. Students reported that weighting of the quizzes, although low, was a motivating factor for completion. The most pronounced differences in academic performance were observed in the mainstream rather than advanced courses, suggesting that the online component mainly benefited students with lower proficiency levels of chemistry. Recommendations regarding the design of the pre-learning materials were proposed to enhance student engagement, encourage the desired behavioural pattern and adoption of a deep learning approach.

In 2010, the state of Tennessee embraced the call to overhaul teacher education and required programs to adopt a residency model within K-12 schools. How exactly this would affect the various methods courses in a teacher education program? This paper provides a description of how two elementary education methods courses have shifted from simulation-style projects to projects that involve working with actual elementary students throughout the semester. This article presents an overview of the new residency style methods courses, along with how major assignments shifted to utilize the extensive time pre-service teachers would spend in the elementary school classroom.

Molecular property prediction has a vast range of applications in the chemical industry. A powerful molecular property prediction model can promote experiments and production processes. The idea behind this degree program lies in the use of transfer learning to predict molecular properties. The project is divided into two parts. The first part is to build and pre-train the model. The model, which is constructed with pure attention-based Transformer Layer, is pre-trained through a Masked Edge Recovery task with large-scale unlabeled data. Then, the performance of this pre- trained model is tested with different molecular property prediction tasks and finally verifies the effectiveness of transfer learning.The results show that after self-supervised pre-training, this model shows its excellent generalization capability. It is possible to be fine-tuned with a short period and performs well in downstream tasks. And the effectiveness of transfer learning is reflected in the experiment as well. The pre-trained model not only shortens the task- specific training time but also obtains better performance and avoids overfitting due to too little training data for molecular property prediction.
Prediktion av molekylers egenskaper har en stor mängd tillämpningar inom kemiindustrin. Kraftfulla metoder för att predicera molekylära egenskaper kan främja vetenskapliga experiment och produktionsprocesser. Ansatsen i detta arbete är att använda överförd inlärning (eng. transfer learning) för att predicera egenskaper hos molekyler. Projektet är indelat i två delar. Den första delen fokuserar på att utveckla och förträna en modell. Modellen består av Transformer-lager med attention- mekanismer och förtränas genom att återställa maskerade kanter i molekylgrafer från storskaliga mängder icke-annoterad data. Efteråt utvärderas prestandan hos den förtränade modellen i en mängd olika uppgifter baserade på prediktion av molekylegenskaper vilket bekräftar fördelen med överförd inlärning.Resultaten visar att modellen efter självövervakad förträning besitter utmärkt förmåga till att generalisera. Den kan finjusteras med liten tidskostnad och presterar väl i specialiserade uppgifter. Effektiviteten hos överförd inlärning visas också i experimenten. Den förtränade modellen förkortar inte bara tiden för uppgifts-specifik inlärning utan uppnår även bättre prestanda och undviker att övertränas på grund otillräckliga mängder data i uppgifter för prediktion av molekylegenskaper.

This research explored the beliefs, knowledge and skills of Bhutanese pre-service mathematics teachers and their preparedness to teach a reformed mathematics curriculum. Data were collected when the participants joined a series of inquiry-oriented, technology-enabled calculus workshops. These data analyses revealed many inconsistencies between the participants’ espoused and enacted beliefs and misalignments between their mathematical knowledge and skills when compared to the intentions of the newly reformed curriculum. From this, implications have been drawn about the preparedness of the soon-to-graduate pre-service mathematics teachers, and recommendations are made regarding ways future pre-service teachers mathematical beliefs, knowledge and skills may be enhanced.

The emergence of ePortfolios is relatively recent in the university sector as a way to engage students in their learning and assessment, and to produce records of their accomplishments. An ePortfolio is an online tool that students can utilise to record, catalogue, retrieve and present reflections and artefacts that support and demonstrate the development of graduate students’ capabilities and professional standards across university courses. The ePortfolio is therefore considered as both process and product. Although ePortfolios show promise as a useful tool and their uptake has grown, they are not yet a mainstream higher education technology. To date, the emphasis has been on investigating their potential to support the multiple purposes of learning, assessment and employability, but less is known about whether and how students engage with ePortfolios in the university setting. This thesis investigates student engagement with an ePortfolio in one university. As the educational designer for the ePortfolio project at the University, I was uniquely positioned as a researching professional to undertake an inquiry into whether students were engaging with the ePortfolio. The participants in this study were a cohort (defined by enrolment in a unit of study) of second and third year education students (n=105) enrolled in a four year Bachelor of Education degree. The students were introduced to the ePortfolio in an introductory lecture and a hands-on workshop in a computer laboratory. They were subsequently required to complete a compulsory assessment task – a critical reflection - using the ePortfolio. Following that, engagement with the ePortfolio was voluntary. A single case study approach arising from an interpretivist paradigm directed the methodological approach and research design for this study. The study investigated the participants’ own accounts of their experiences with the ePortfolio, including how and when they engaged with the ePortfolio and the factors that impacted on their engagement. Data collection methods consisted of an attitude survey, student interviews, document collection, a researcher reflective journal and researcher observations. The findings of the study show that, while the students were encouraged to use the ePortfolio as a learning and employability tool, most students ultimately chose to disengage after completing the assessment task. Only six of the forty-five students (13%) who completed the research survey had used the ePortfolio in a sustained manner. The data obtained from the students during this research has provided insight into reasons why they disengaged from the ePortfolio. The findings add to the understandings and descriptions of student engagement with technology, and more broadly, advance the understanding of ePortfolios. These findings also contribute to the interdisciplinary field of technology implementation. There are three key outcomes from this study, a model of student engagement with technology, a set of criteria for the design of an ePortfolio, and a set of recommendations for effective practice for those implementing ePortfolios. The first, the Model of Student Engagement with Technology (MSET) (Version 2) explored student engagement with technology by highlighting key engagement decision points for students The model was initially conceptualised by building on work of previous research (Version 1), however, following data analysis a new model emerged, MSET (Version 2). The engagement decision points were identified as: • Prior Knowledge and Experience, leading to imagined usefulness and imagined ease of use; • Initial Supported Engagement, leading to supported experience of usefulness and supported ease of use; • Initial Independent Engagement, leading to actual experience of independent usefulness and actual ease of use; and • Ongoing Independent Engagement, leading to ongoing experience of usefulness and ongoing ease of use. The Model of Student Engagement with Technology (MSET) goes beyond numerical figures of usage to demonstrate student engagement with an ePortfolio. The explanatory power of the model is based on the identification of the types of decisions that students make and when they make them during the engagement process. This model presents a greater depth of understanding student engagement than was previously available and has implications for the direction and timing of future implementation, and academic and student development activities. The second key outcome from this study is a set of criteria for the re-conceptualisation of the University ePortfolio. The knowledge gained from this research has resulted in a new set of design criteria that focus on the student actions of writing reflections and adding artefacts. The process of using the ePortfolio is reconceptualised in terms of privileging student learning over administrative compliance. The focus of the ePortfolio is that the writing of critical reflections is the key function, not the selection of capabilities. The third key outcome from this research consists of five recommendations for university practice that have arisen from this study. They are that, sustainable implementation is more often achieved through small steps building on one another; that a clear definition of the purpose of an ePortfolio is crucial for students and staff; that ePortfolio pedagogy should be the driving force not the technology; that the merit of the ePortfolio is fostered in students and staff; and finally, that supporting delayed task performance is crucial. Students do not adopt an ePortfolio just because it is provided. While students must accept responsibility for their own engagement with the ePortfolio, the institution has to accept responsibility for providing the environment, and technical and pedagogical support to foster engagement. Ultimately, an ePortfolio should be considered as a joint venture between student and institution where strong returns on investment can be realised by both. It is acknowledged that the current implementation strategies for the ePortfolio are just the beginning of a much longer process. The real rewards for students, academics and the university lie in the future.

The purpose of this qualitative case study was to explore the implementation of a Professional Learning Community (PLC) as a professional development model effective in altering teachers‘ perceptions of their knowledge and skill in teaching developmental writing in grades K-3. This research is necessary to examine how offering teachers collaborative support needed for understanding and implementing research-based best practice approaches to teach developmental writing strengthens the quality of instructional practice necessary to meet rigorous standards being imparted from Common Core Standards. Through the development of a PLC, teachers have an opportunity for collaborating within the school building, which provides optimal environment for professional development (Lindeman, 1926; Lumpe, 2007). Analysis taken from the pre and post-survey information included the teachers‘ beliefs and understanding of writing development, current use of instruction time for writing, and questions or concerns teachers have about teaching writing. Weekly PLC meetings using a protocol format offered teachers an opportunity to discuss personal experiences with writing instruction and to share any anchor charts, student work, or anecdotal records exemplifying the strategy of focus. Videotaping and reflective journaling collected during the six PLC sessions were transcribed and coded using predetermined and emerging themes within and across each measure. Presentation materials collected as data documentation of the experience aided in validation of the research. Major themes emerged under the code headed as management with sub-codes of planning and classroom management presenting the strongest focus. Major themes also emerged under the code headed as instruction. The strongest areas of focus under the instruction code included subcoded areas conferencing, minilessons, and teaching strategies.

Energy efficiency of district heating systems is of great interest to energy stakeholders. However, it is not uncommon that district heating systems fail to achieve the expected performance due to inappropriate operations. Night setback is one control strategy, which has been proved to be not a suitable setting for well-insulated modern buildings in terms of both economic and energy efficiency. Therefore, identification of a night setback control is vital to district heating companies to smoothly manage their heat energy distribution to their customers. This study is motivated to automate this identification process. The method used in this thesis is a Convolutional Neural Network(CNN) approach using the concept of transfer learning. 133 substations in Oslo are used in this case study to design a machine learning model that can identify a substation as night setback or non-night setback series. The results show that the proposed method can classify the substations with approximately 97% accuracy and 91% F1-score. This shows that the proposed method has a high potential to be deployed and used in practice to identify a night setback control in district heating substations.

During the lifetime of a satellite malfunctions may occur. Unexpected behaviour are monitored using sensors all over the satellite. The telemetry values are then sent to Earth and analysed seeking for anomalies. These anomalies could be detected by humans, but this is considerably expensive. To lower the costs, machine learning techniques can be applied. In this research many diferent machine learning techniques are tested and compared using satellite telemetry data provided by OHB System AG. The fact that the anomalies are collective, together with some data properties, is exploited to improve the performances of the machine learning algorithms. Since the data comes from a real spacecraft, it presents some defects. The data covers in fact a small time-lapse and does not present critical anomalies due to the spacecraft healthiness. Some steps are then taken to improve the evaluation of the algorithms.

In software development projects, bug triage consists mainly of assigning bug reports to software developers or teams (depending on the project). The partial or total automation of this task would have a positive economic impact on many software projects. This thesis introduces a systematic four-step method to find some of the best configurations of several machine learning algorithms intending to solve the automatic bug assignment problem. These four steps are respectively used to select a combination of pre-processing techniques, a bug report representation, a potential feature selection technique and to tune several classifiers. The aforementioned method has been applied on three software projects: 66 066 bug reports of a proprietary project, 24 450 bug reports of Eclipse JDT and 30 358 bug reports of Mozilla Firefox. 619 configurations have been applied and compared on each of these three projects. In production, using the approach introduced in this work on the bug reports of the proprietary project would have increased the accuracy by up to 16.64 percentage points.

In response to social trends whereby children are spending less time outside, school administrators have developed certified Nature Explore Outdoor Classrooms (NEOCs) intentionally designed to support whole-child learning within a natural environment. Despite the documented benefits of nature-based education, the literature and NEOC sites report challenges in facilitating this type of space. The purpose of this study was to investigate what prevents teachers in a certified NEOC from facilitating student/teacher engagement with the natural outdoor environment. Kolb's, Piaget's, and Vygotsky's theories of constructivism served as the study's framework to explore the problem from the teachers' perspectives. A qualitative case study was used to gain insight into the potential barriers to facilitating a NEOC. Eight teachers were recruited using purposeful sampling. Participant criteria included (a) >18 years of age, (b) >3 years early childhood teaching experience, (c) >1 year experience in selected NEOC, (d) prior NEOC training, and (e) willingness to share experiences. Data collection included classroom observation, individual interviewing, and review of relevant documents. All data were analyzed using comparative and inductive analysis and coded into 5 emergent themes. Identified barriers included teacher involvement, rules and regulations, volunteers, materials, and weather. By creating a 3-day professional development program that supports the benefits of nature-based learning environments and introduces strategies to overcome identified barriers, this study may promote positive social change in nature-based education. Children, families, and communities may expand their nature-based knowledge and interaction skills to pass to future generations.

Young novice drivers - that is, drivers aged 16-25 years who are relatively inexperienced in driving on the road and have a novice (Learner, Provisional) driver's licence - have been overrepresented in car crash, injury and fatality statistics around the world for decades. There are numerous persistent characteristics evident in young novice driver crashes, fatalities and offences, including variables relating to the young driver themselves, broader social influences which include their passengers, the car they drive, and when and how they drive, and their risky driving behaviour in particular. Moreover, there are a range of psychosocial factors influencing the behaviour of young novice drivers, including the social influences of parents and peers, and person-related factors such as age-related factors, attitudes, and sensation seeking. Historically, a range of approaches have been developed to manage the risky driving behaviour of young novice drivers. Traditional measures predominantly relying upon education have had limited success in regulating the risky driving behaviour of the young novice driver. In contrast, interventions such as graduated driver licensing (GDL) which acknowledges young novice drivers' limitations - principally pertaining to their chronological and developmental age, and their driving inexperience - have shown to be effective in ameliorating this pervasive public health problem. In practice, GDL is a risk management tool that is designed to reduce driving at risky times (e.g., at night) or in risky driving conditions (e.g., with passengers), while still enabling novice drivers to obtain experience. In this regard, the GDL program in Queensland, Australia, was considerably enhanced in July 2007, and major additions to the program include mandated Learner practice of 100 hours recorded in a logbook, and passenger limits during night driving in the Provisional phase. Road safety researchers have also continued to consider the influential role played by the young driver's psychosocial characteristics, including psychological traits and states. In addition, whilst the majority of road safety user research is epidemiological in nature, contemporary road safety research is increasingly applying psychological and criminological theories. Importantly, such theories not only can guide young novice driver research, they can also inform the development and evaluation of countermeasures targeting their risky driving behaviour. The research is thus designed to explore the self-reported behaviours - and the personal, psychosocial, and structural influences upon the behaviours - of young novice drivers This thesis incorporates three stages of predominantly quantitative research to undertake a comprehensive investigation of the risky driving behaviour of young novices. Risky driving behaviour increases the likelihood of the young novice driver being involved in a crash which may harm themselves or other road users, and deliberate risky driving such as driving in excess of the posted speed limits is the focus of the program of research. The extant literature examining the nature of the risky behaviour of the young novice driver - and the contributing factors for this behaviour - while comprehensive, has not led to the development of a reliable instrument designed specifically to measure the risky behaviour of the young novice driver. Therefore the development and application of such a tool (the Behaviour of Young Novice Drivers Scale, or BYNDS) was foremost in the program of research. In addition to describing the driving behaviours of the young novice, a central theme of this program of research was identifying, describing, and quantifying personal, behavioural, and environmental influences upon young novice driver risky behaviour. Accordingly the 11 papers developed from the three stages of research which comprise this thesis are framed within Bandura's reciprocal determinism model which explicitly considers the reciprocal relationship between the environment, the person, and their behaviour. Stage One comprised the foundation research and operationalised quantitative and qualitative methodologies to finalise the instrument used in Stages Two and Three. The first part of Stage One involved an online survey which was completed by 761 young novice drivers who attended tertiary education institutions across Queensland. A reliable instrument for measuring the risky driving behaviour of young novices was developed (the BYNDS) and is currently being operationalised in young novice driver research in progress at the Centre for Injury Research and Prevention in Philadelphia, USA. In addition, regression analyses revealed that psychological distress influenced risky driving behaviour, and the differential influence of depression, anxiety, sensitivity to punishments and rewards, and sensation seeking propensity were explored. Path model analyses revealed that punishment sensitivity was mediated by anxiety and depression; and the influence of depression, anxiety, reward sensitivity and sensation seeking propensity were moderated by the gender of the driver. Specifically, for males, sensation seeking propensity, depression, and reward sensitivity were predictive of self-reported risky driving, whilst for females anxiety was also influential. In the second part of Stage One, 21 young novice drivers participated in individual and small group interviews. The normative influences of parents, peers, and the Police were explicated. Content analysis supported four themes of influence through punishments, rewards, and the behaviours and attitudes of parents and friends. The Police were also influential upon the risky driving behaviour of young novices. The findings of both parts of Stage One informed the research of Stage Two. Stage Two was a comprehensive investigation of the pre-Licence and Learner experiences, attitudes, and behaviours, of young novice drivers. In this stage, 1170 young novice drivers from across Queensland completed an online or paper survey exploring their experiences, behaviours and attitudes as a pre- and Learner driver. The majority of novices did not drive before they were licensed (pre-Licence driving) or as an unsupervised Learner, submitted accurate logbooks, intended to follow the road rules as a Provisional driver, and reported practicing predominantly at the end of the Learner period. The experience of Learners in the enhanced-GDL program were also examined and compared to those of Learner drivers who progressed through the former-GDL program (data collected previously by Bates, Watson, & King, 2009a). Importantly, current-GDL Learners reported significantly more driving practice and a longer Learner period, less difficulty obtaining practice, and less offence detection and crash involvement than Learners in the former-GDL program. The findings of Stage Two informed the research of Stage Three. Stage Three was a comprehensive exploration of the driving experiences, attitudes and behaviours of young novice drivers during their first six months of Provisional 1 licensure. In this stage, 390 of the 1170 young novice drivers from Stage Two completed another survey, and data collected during Stages Two and Three allowed a longitudinal investigation of self-reported risky driving behaviours, such as GDL-specific and general road rule compliance; risky behaviour such as pre-Licence driving, crash involvement and offence detection; and vehicle ownership, paying attention to Police presence, and punishment avoidance. Whilst the majority of Learner and Provisional drivers reported compliance with GDL-specific and general road rules, 33% of Learners and 50% of Provisional drivers reported speeding by 10-20 km/hr at least occasionally. Twelve percent of Learner drivers reported pre-Licence driving, and these drivers were significantly more risky as Learner and Provisional drivers. Ten percent of males and females reported being involved in a crash, and 10% of females and 18% of males had been detected for an offence, within the first six months of independent driving. Additionally, 75% of young novice drivers reported owning their own car within six months of gaining their Provisional driver's licence. Vehicle owners reported significantly shorter Learner periods and more risky driving exposure as a Provisional driver. Paying attention to Police presence on the roads appeared normative for young novice drivers: 91% of Learners and 72% of Provisional drivers reported paying attention. Provisional drivers also reported they actively avoided the Police: 25% of males and 13% of females; 23% of rural drivers and 15% of urban drivers. Stage Three also allowed the refinement of the risky behaviour measurement tool (BYNDS) created in Stage One; the original reliable 44-item instrument was refined to a similarly reliable 36-item instrument. A longitudinal exploration of the influence of anxiety, depression, sensation seeking propensity and reward sensitivity upon the risky behaviour of the Provisional driver was also undertaken using data collected in Stages Two and Three. Consistent with the research of Stage One, structural equation modeling revealed anxiety, reward sensitivity and sensation seeking propensity predicted self-reported risky driving behaviour. Again, gender was a moderator, with only reward sensitivity predicting risky driving for males. A measurement model of Akers' social learning theory (SLT) was developed containing six subscales operationalising the four constructs of differential association, imitation, personal attitudes, and differential reinforcement, and the influence of parents and peers was captured within the items in a number of these constructs. Analyses exploring the nature and extent of the psychosocial influences of personal characteristics (step 1), Akers' SLT (step 2), and elements of the prototype/willingness model (PWM) (step 3) upon self-reported speeding by the Provisional driver in a hierarchical multiple regression model found the following significant predictors: gender (male), car ownership (own car), reward sensitivity (greater sensitivity), depression (greater depression), personal attitudes (more risky attitudes), and speeding (more speeding) as a Learner. The research findings have considerable implications for road safety researchers, policy-makers, mental health professionals and medical practitioners alike. A broad range of issues need to be considered when developing, implementing and evaluating interventions for both the intentional and unintentional risky driving behaviours of interest. While a variety of interventions have been historically utilised, including education, enforcement, rehabilitation and incentives, caution is warranted. A multi-faceted approach to improving novice road safety is more likely to be effective, and new and existing countermeasures should capitalise on the potential of parents, peers and Police to be a positive influence upon the risky behaviour of young novice drivers. However, the efficacy of some interventions remains undetermined at this time. Notwithstanding this caveat, countermeasures such as augmenting and strengthening Queensland's GDL program and targeting parents and adolescents particularly warrant further attention. The findings of the research program suggest that Queensland's current-GDL can be strengthened by increasing compliance of young novice drivers with existing conditions and restrictions. The rates of speeding reported by the young Learner driver are particularly alarming for a number of reasons. The Learner is inexperienced in driving, and travelling in excess of speed limits places them at greater risk as they are also inexperienced in detecting and responding appropriately to driving hazards. In addition, the Learner period should provide the foundation for a safe lifetime driving career, enabling the development and reinforcement of non-risky driving habits. Learners who sped reported speeding by greater margins, and at greater frequencies, when they were able to drive independently. Other strategies could also be considered to enhance Queensland's GDL program, addressing both the pre-Licence adolescent and their parents. Options that warrant further investigation to determine their likely effectiveness include screening and treatment of novice drivers by mental health professionals and/or medical practitioners; and general social skills training. Considering the self-reported pre-licence driving of the young novice driver, targeted education of parents may need to occur before their child obtains a Learner licence. It is noteworthy that those participants who reported risky driving during the Learner phase also were more likely to report risky driving behaviour during the Provisional phase; therefore it appears vital that the development of safe driving habits is encouraged from the beginning of the novice period. General education of parents and young novice drivers should inform them of the considerably-increased likelihood of risky driving behaviour, crashes and offences associated with having unlimited access to a vehicle in the early stages of intermediate licensure. Importantly, parents frequently purchase the car that is used by the Provisional driver, who typically lives at home with their parents, and therefore parents are ideally positioned to monitor the journeys of their young novice driver during this early stage of independent driving. Parents are pivotal in the development of their driving child: they are models who are imitated and are sources of attitudes, expectancies, rewards and punishments; and they provide the most driving instruction for the Learner. High rates of self-reported speeding by Learners suggests that GDL programs specifically consider the nature of supervision during the Learner period, encouraging supervisors to be vigilant to compliance with general and GDL-specific road rules, and especially driving in excess of speed limit. Attitudes towards driving are formed before the adolescent reaches the age when they can be legally licensed. Young novice drivers with risky personal attitudes towards driving reported more risky driving behaviour, suggesting that countermeasures should target such attitudes and that such interventions might be implemented before the adolescent is licensed. The risky behaviours and attitudes of friends were also found to be influential, and given that young novice drivers tend to carry their friends as their passengers, a group intervention such as provided in a school class context may prove more effective. Social skills interventions that encourage the novice to resist the negative influences of their friends and their peer passengers, and to not imitate the risky driving behaviour of their friends, may also be effective. The punishments and rewards anticipated from and administered by friends were also found to influence the self-reported risky behaviour of the young novice driver; therefore young persons could be encouraged to sanction the risky, and to reward the non-risky, driving of their novice friends. Adolescent health programs and related initiatives need to more specifically consider the risks associated with driving. Young novice drivers are also adolescents, a developmental period associated with depression and anxiety. Depression, anxiety, and sensation seeking propensity were found to be predictive of risky driving; therefore interventions targeting psychological distress, whilst discouraging the expression of sensation seeking propensity whilst driving, warrant development and trialing. In addition, given that reward sensitivity was also predictive, a scheme which rewards novice drivers for safe driving behaviour - rather than rewarding the novice through emotional and instrumental rewards for risky driving behaviour - requires further investigation. The Police were also influential in the risky driving behaviour of young novices. Young novice drivers who had been detected for an offence, and then avoided punishment, reacted differentially, with some drivers appearing to become less risky after the encounter, whilst for others their risky behaviour appeared to be reinforced and therefore was more likely to be performed again. Such drivers saw t

Comparative study between the performance of Convolutional Networks using pretrained models and statistical generative models on tasks of image classification in semi-supervised enviroments.Study of multiple ensembles using these techniques and generated data from estimated pdfs.Pretrained Convents, LDA, pLSA, Fisher Vectors, Sparse-coded SPMs, TSVMs being the key models worked upon.

Comparative study between the performance of Convolutional Networks using pretrained models and statistical generative models on tasks of image classification in semi-supervised enviroments.Study of multiple ensembles using these techniques and generated data from estimated pdfs.Pretrained Convents, LDA, pLSA, Fisher Vectors, Sparse-coded SPMs, TSVMs being the key models worked upon.

Student progression is historically dominated by indicators of academic success, denoted by passing subjects and successfully completing qualifications. More recently this university management view of progression has been linked to institutional audit measures with these providing secondary data archives for research into progression. This thesis provides a student perspective on progression. To do so it necessarily encompasses wider student goals and aspirations that extend beyond the university to include career and professional growth. The research uses a sequential mixed methods approach with three main studies (archive, survey and interview) each aligned with a research objective. The first study in the series is an archive study of 1547 university Built Environment student records that were originally collected as part of the requirement for the university to report its work for government audit. The second study was a self-report structured online survey that collected data from a total of 295 students. The student sample was made up of 154 Built Environment students and 141 Engineering students. 97, all Built Environment students, were distance education students and the other 198 students were campus-based students. The final study, the in-depth interview study, used one-on-one interactive conversational interviews with students, graduates, academics, employers and executives from professional organisations. In total 40 interviews were undertaken. A number of influences were found on progression as seen from a student perspective. These include the importance of work-readiness, cadetships and the alignment of qualifications with industry needs and career growth opportunities. The student user perspective also brought into focus a number of the pedagogical assumptions held about and held by students in the Built Environment.

碩士
中國文化大學
國際企業管理學系碩士在職專班
104
Lifestyles nowadays have undergone a tremendous change with the rapid development of wireless network as well as mobile technology, which also influenced the humanity’s way of learning. Learning is no longer confined in an existing classroom; instead, mobile learning which possesses mobility, convenience and expediency, have developed into a new approach of digital learning in the current age. While enterprises constantly pursue sustainable management, rapid learning and knowledge innovation, as a matter of fact, are the only crucial factors for an enterprise to establish long-term competitive advantage. Due to this reason, making use of mobile learning system is of immense importance to the need of rapid learning in enterprises. Trainings in direct selling industry, likewise, surmount the traditional mode by applying mobile learning platforms, offering distributors an immediate, convenient approach of learning. As yet, the objects of numerous literatures on studying the behavior of using mobile learning are mostly students. There are relatively less studies on the usage of mobile learning in enterprises, not to mention in direct selling. Hence, based on the unified theory of acceptance and use of technology (UTAUT), the purpose of this study is to investigate and understand the usage of distributors using the mobile learning system from direct selling industry which already have the system built in for two years. vi A total of 663 effective samples were collected from the distributors who have already used the mobile learning system built by a direct selling industry. Through the analysis and research by SPSS 22, the results are as follows: 1. Performance expectancy has a significant and positive effect on the use behavior and intention of the users of mobile learning system. 2. Effort expectancy has a significant and positive effect on the use behavior and intention of the users of mobile learning system. 3. Social influence has a significant and positive effect on the use behavior and intention of the users of mobile learning system. 4. Facilitating condition has a significant and positive effect on the use behavior of the users of mobile learning system. 5. Behavior intention has a significant and positive effect on the use behavior of the users of mobile learning system. In an attempt to increase users’ intention to use mobile learning system, the user interface of the platform ought to be easier to manipulate, and the course contents brief and simple. Meanwhile, we should let users understand the benefits of learning through this system, via the reviews from the users who found it effective. In this way, we are able to improve the simplicity of mobile learning system as well as working efficiency, and further increase the usage intention and rate.

The term e-portfolio refers to a portfolio in electronic format that allows users to collect evidence of learning in several media types (e.g., audio, video, text, and graphics) and to organise these using hypertext links (Barrett, 2001). E-portfolios have been introduced into teacher education programs internationally to help pre-service teachers (PST) build records of their learning and reflections, and allow them to assemble collections of evidence of their achievements against graduate standards. These e-portfolios may function as digital CVs; and also support lifelong learning after graduation (Oakley, Pegrum, & Johnston, 2014). Through investigating the experience of e-portfolio use by PSTs, this thesis provides significant evidence about the high quality implementation of e-portfolios in higher education. The thesis explores the reasons behind the participants’ success in an e-portfolio-based unit. In particular, the research explores the reasons why a number of the participants were more successful than others when using e-portfolios. This is the first research which has examined PSTs perspectives on e-portfolio-based learning within constructivism, students’ approach to learning (SAL), the 3P model (presage, process, and product) of learning, and self-regulated learning (SRL). An e-portfolio-based unit in the Faculty of Education in an Australian University was investigated using a mixed methods research design to analyse the data gathered through conducting pre-unit and post-unit interviews. The qualitative analysis examines the participants’ conceptions of e-portfolios, their perceptions of the teaching and learning context, and the effect of these on their approaches to learning and their learning outcomes. A questionnaire was distributed at week 11 to measure how they conceived e-portfolios, how they perceived the quality of the teaching, the clarity of the goals, and the appropriateness of the assessment and workload. This research showed that there was variation in the academic achievements of the PSTs when using e-portfolios and the results of the analysis confirmed that the learning outcomes at the surface or deep approach to learning were affected by the participants’ conceptions of the e-portfolios, their perceived role, and the perceptions of their lecturers’ role. In particular, their experience in the course depended on their perception of good teaching, clarity of their goals, and appropriate workload and assessment in the unit. Therefore, these factors seemed to be significantly related to what they did, and the strategies they used when using the e-portfolio. The implications of the results of this thesis are relevant for educators responsible for designing new e-portfolio-based units or courses, and improving the teaching and learning outcomes of existing e-portfolio-based learning.

碩士
國立臺灣師範大學
體育學系
103
The purpose of the study was to investigate how pre-service PE teachers perceive and implement the Cooperative Learning Instruction Model (CLIM) in physical education during their 5 weeks teaching practicum in a secondary school. Four pre-service PE teachers were participants of the study. Each pre-service PE teachers had 2 lessons per week (total 9 lessons). Focus-group interviews, Semi-structured interviews, class observation, and reflection journals were used to collect data. The interview data were transcribed in order to further comparison, integration and analysis. The interview data were transcribed for further comparison, integration and analysis. Results were showed that (1) before the implementation of CLIM: pre-service PE Teachers hold suspect attitudes toward this instructional model; (2) During the implementation of CLIM: pre-service PE teachers could follow the instructional steps, however, But for students learning motor skills more dissatisfied, Although the perception of the development of social skills and thinking skills, but there was also a need for developing their observation, interpretation and teaching decision competencies; (3) after the implementation of CLIM: pre-service PE teachers recognized CLIM as a effective way for learning, though the grouping skills and quick decision ability while teaching would be need for developing .

Le développement professionnel des enseignants a longtemps été identifié comme essentiel à la réussite de l'adoption des TIC en éducation (BECTA, 2009). Des programmes de formation efficaces sur les TIC pour le développement professionnel des futurs enseignants devraient veiller à ce que, une fois diplômés, les enseignants aient les compétences et les connaissances nécessaires pour utiliser efficacement les TIC dans les salles de classe non seulement en raison de la nécessité pour les enfants de développer des compétences qui leur permettront de bien se débrouiller dans la société moderne (UNESCO, 2011; Dede, 2014), mais aussi, en raison de la valeur potentielle de ces technologies comme outils d'apprentissage (Gill and Dalgarno, 2008). Les TIC sont donc, devenues des incontournables pour les enseignants et les apprenants dans le contexte de la société du savoir. Cependant, lorsqu’on regarde leurs usages et leurs impacts on n’est pas satisfait après tout ce qui a été investi en termes de formation, argent, équipement, etc. Une des causes semble être la formation des enseignants, notamment dans son contenu et dans les stratégies de formation adoptée (Villeneuve, et al. 2012). Nous avons développé une intervention de formation pour rendre opérationnel le modèle IntersTICES (Peraya and Viens, 2005) et aider les formateurs d'enseignants intégrer les TIC dans leur pratique d'enseignement. Cette opérationnalisation impliquant les enseignants travaillant dans le programme de formation initiale des maîtres à l'Université de Montréal, a mis en perspective l'importance de la culture e-learning des formateurs d'enseignants, de l’accompagnement, et des interventions de suivi pour les activités, y compris l'utilisation pédagogique des TIC. La recherche a porté sur l'analyse de l'impact de l'intervention de formation sur la culture e-learning de participants, ainsi que sur leur intention d'adopter et d'utiliser les TIC dans leurs cours. Les résultats suggèrent que l’opérationnalisation du modèle IntersTICES via une intervention de formation de type IntersTICES, peut fournir les formateurs des formateurs une occasion de réflexion et de sensibilisation sur leurs représentations personnelles concernant tous les aspects de leur culture e-learning. De plus, cette opérationnalisation a aidé les formateurs des formateurs à prendre conscience de l'impact que leur culture e-learning a sur leur pratique tout au long et à n’importe quel stade du développement et de la mise en œuvre de leur activité choisie intégrant les TIC.
In today's knowledge society, ICT has become essential for teachers and learners (BECTA, 2009) not only because of the need for children to acquire and develop skills that will help them grow as collaborative, problem-solving, creative learners (UNESCO, 2011; Dede, 2014), but also because of the potential value of such technologies as tools for learning (Gill and Dalgarno, 2008). However, when looking at the current educational ICT uses reported by research conducted with teachers and students, (Karsenti and Lira, 2010; Villeneuve et al., 2012) and the low impact observed on learning - despite the substantial investment in equipment and material -we can only question the possible reasons for such a situation. One of the causes identified in the literature is teacher training, especially in its content and the training strategies adopted by teacher educators (Angeli, 2009; Enochson, and Rizza, 2009). We developed a training intervention to operationalize the IntersTICES model (Peraya and Viens 2005) and help teacher trainers integrate ICT in their teaching practice. This operationalization involving teachers working in the teacher training program at Université de Montreal, put into perspective the importance of teacher trainers’ e-learning culture, personal support and follow-up interventions for activities including pedagogical use of ICT. The research focused on analyzing the impact of the training intervention on participants’ e-learning culture, as well as on their intention to adopt and use ICT in their courses. Findings suggest that the operationalization of the IntersTICES model via an interactive training intervention, can provide teacher trainers with an opportunity for reflection and awareness about their personal representations regarding every aspect of their e-learning culture. Furthermore, it helped teacher trainers becoming aware of the impact their e-learning culture has on their practice throughout and at any stage of the development and implementation of their chosen activity integrating ICT.

Dissertation presented as the partial requirement for obtaining a Master's degree in Information Management, specialization in Knowledge Management and Business Intelligence
Existe um pré-pagamento quando ocorre um reembolso antecipado de um empréstimo por parte do tomador, i.e., o tomador paga mais que o montante contratual acordado. Tal pode ocorrer como parte do principal em dívida (reembolso parcial) ou o valor total do principal em dívida (reembolso total). Do ponto de vista de um banco, o estudo do reembolso antecipado - seja total ou parcial - é importante, pois resulta numa mudança nos fluxos de caixa calendarizados. Em particular, há uma diminuição nos fluxos de caixa futuros resultantes de um evento futuro desconhecido. Assim, o principal objetivo deste estudo é a modelação dos eventos de pré-pagamento no crédito à habitação de um grande banco português, através de uma abordagem de machine learning, avaliando o seu desempenho através da utilização de técnicas como a Area Under the Receiver Operating Characteristic Curve (ROC), o gain or lift e Kolmogorov-Smirnov. Tal permite o estudo do fenómeno das amortizações antecipadas (ou pré-pagamentos) no mercado Português, utilizando dados reais, e através de modelos de machine learning. Uma vez que foram utilizados dados reais, a primeira parte deste estudo prendeu-se com o préprocessamento dos dados, de modo a garantir que os modelos não incluíam ruído e problemas de qualidade de dados. A segunda parte prendeu-se com a computação dos modelos de machine learning, testando modelos de artificial neural network e random forest, com a comparação da performance destes através de métricas como o ROC, gain or lift e Kolmogorov-Smirnov. Os resultados obtidos revelam que os modelos de pré-pagamento total e parcial apresentam bom desempenho nas três métricas de desempenho analisadas. Ambos os modelos apresentam resultados positivos e demonstram que os modelos apresentam bons resultados preditivos e capacidade discriminatória, sendo o modelo de amortização parcial superior ao modelo de amortização total, com uma diferença que, embora não muito grande, merece destaque. Este estudo é particularmente relevante dada a sua análise num banco português, e a aplicação de modelos de machine learning na modelação de pré-pagamento, para os quais os estudos são escassos. Por outro lado, têm recentemente ocorrido esforços (por parte do banco onde o estudo se encontra incluído) para a atualização dos modelos tradicionais atualmente em vigor.
There is a loan prepayment when there is an early repayment of a loan from the borrower, i.e. the borrower pays more than the contractual amount due. The repayment may be part of the outstanding principal (partial repayment) or the total principal outstanding (full repayment). From a Bank’s perspective, the study of early repayment – be it full or partial – is relevant as they result in a change in the schedule cash flows. In particular, there is a decrease in the future cash flows resulting from an unknown future event. Hence, the primary purpose of this study is the modelling of the prepayment events in the mortgage loans of a large Portuguese bank, through a machine learning approach, assessing its performance through the use of techniques such as the Area Under the Receiver Operating Characteristic Curve (ROC), the Gain or Lift, and Kolmogorov-Smirnov statistic. This allows for the test of the prepayment phenomena in the Portuguese reality, using real Bank data, and through the use of machine learning models. As there was a use of real-life data, the first part of this study implied the pre-processing of the data, to ensure that the noise and data quality problems were not part of the models. The second stage implied the computation of the machine learning models, which occurred through the testing of Artificial Neural Network and Random Forest models, with the comparison of its performance using the ROC, Gain or Lift and Kolmogorov-Smirnov statistic. The results obtained reveal that both the total and partial prepayment models perform well in all the three performance metrics analysed. Both models present positive results and demonstrate that the models have good predictive results and discriminatory capacity. The partial repayment model is superior to the full repayment model, with a difference that is worthy of mention although not very large. This study is particularly relevant given its analysis in a Portuguese bank and the application of machine learning models in modelling prepayment, for which studies are scarce. Furthermore, there have been occurring efforts (in the bank where this study is framed) to update the traditional models currently in force.

Sport coaches play a crucial role in contributing to individual and team performance. Unlike the scientific study of athlete development, scant attention has been paid to the intricacies of coach development across the full spectrum of coaching levels. In the current research the defining stage development characteristics of A, B and C license soccer (hereafter referred to as football) coaches in Australia and Malaysia, were interviewed about their development journey, especially dating back to their first involvement in football through to the present.

There are numerous problems associated with the teaching of Euclidean geometry at secondary schools today. Students do not see the necessity of proving results which have been obtained intuitively. They do not comprehend that the validity of a deduction is independent of the 'truth' of the initial assumptions. They do not realise that they cannot reason from diagrams, because these may be misleading or inaccurate. Most importantly, they do not understand that Euclidean geometry is a particular interpretation of physical space and that there are alternative, equally valid interpretations. A possible means of addressing the above problems is tbe introduction of nonEuclidean geometry at school level. It is imperative to identify those students who have the pre-requisite knowledge and skills. A number of interesting teaching strategies, such as debates, discussions, investigations, and oral and written presentations, can be used to introduce and develop the content matter.
Mathematics Education
M. Sc. (Mathematics)

Summaries in Afrikaans and English
Text in Afrikaans
With compulsory education in the New South· Africa, an increase in the number of school beginners can be expected, together with the probability of large gaps in their readiness for school. In £his study, materials have been developed together with a structure for the organisation of teaching_~!?.r trainers/subject advisers, teachers and pupils, to ensure effective education for school beginners. The "Battery for the Assessment of Bridging period Yields" (BABY) has been developed as a valid, realiable and usable school readiness battery of tests for teachers as unregistered test users. The Pre-Test for division into homogeneous ability groups is applied shortly after entrance to school. The Battery of Diagnostic Tests ensures continuous evaluation of pupils during the school readiness programme. The Post-Test is applied at the end of the first term. In this dynamic interaction between tester and pupil in an instruction-test-instruction cycle, the pupils' readiness to progress with formal instruction is continuously evaluated.
Met verpligte onderwys in die Nuwe Suid-Afrika, kan 'n toename in die getal swart skoolbeginners ve:wag word, asook dat daar in hulle skoolvoorbereiding groot leemtes kan voorkom. In die onderhawige studie is hulpmiddels en 'n struktuur vir onderwysorganisasie vir opleiers/vakadviseurs, onderwyseresse en leerlinge ontwikkel om aan skoolbeginners effektiewe onderwys te verseker. Die "Battery for the Assessment of Bridging period Yields" (BABY), 'n geldige, betroubare en bruikbare skoolgereedheidstoetsbattery vir onderwyseresse as ongeregistreerde toetsgebruikers, is ontwikkel. Die Voortoets vtr indeling in homogene bevoegdheidsgroepe word kort na skooltoetrede toegepas. Die Diagnostiese Toetsbattery verseker deurlopende evaluering van leerlinge gedurende die skoolgereedmakingsprogram. Die Natoets word aan die einde van die eerste kwartaal toegepas. In hierdie dinamiese interaksie tussen toetser en leerling in 'n onderrig-toets-onderrig siklus, word leerlinge se gereedheid om met formele onderrig voort te gaan, deurlopend geevalueer.
Psychology og Education
M. Ed. (Psychology of Education)