Ética, estética e experiência formativa: possibilidade da Bildung no presente

Pós-graduação em Educação - FFC

O ofuscamento da infância no brilho das telas: relações entre teoria crítica, educação e sociedade

Pós-graduação em Educação Escolar - FCLAR

Effective employment-based training models for childcare workers

Childcare workers play a significant role in the learning and development of children in their care. This has major implications for the training of workers. Under new reforms of the childcare industry the Australian government now requires all workers to obtain qualifications from a vocational education and training provider (eg. Technical and Further Education) or university. Effective models of employment-based training are critical to provide training to highly competent workers. This paper presents findings from a study that examined current and emerging models of employment-based training in the childcare sector, particularly at the Diploma level. Semi-structured interviews were conducted with a sample of 16 participants who represented childcare directors, employers, and workers located in childcare services in urban, regional and remote locations in the State of Queensland. The study proposes a ‘best-fit’ employment-based training approach that is characterised by a compendium of five models instead of a ‘one size fits all’. Issues with successful implementation of the EBT models are also discussed

Learning the Kernel Matrix with Semi-Definite Programming

Kernel-based learning algorithms work by embedding the data into a Euclidean space, and then searching for linear relations among the embedded data points. The embedding is performed implicitly, by specifying the inner products between each pair of points in the embedding space. This information is contained in the so-called kernel matrix, a symmetric and positive definite matrix that encodes the relative positions of all points. Specifying this matrix amounts to specifying the geometry of the embedding space and inducing a notion of similarity in the input space -- classical model selection problems in machine learning. In this paper we show how the kernel matrix can be learned from data via semi-definite programming (SDP) techniques. When applied to a kernel matrix associated with both training and test data this gives a powerful transductive algorithm -- using the labelled part of the data one can learn an embedding also for the unlabelled part. The similarity between test points is inferred from training points and their labels. Importantly, these learning problems are convex, so we obtain a method for learning both the model class and the function without local minima. Furthermore, this approach leads directly to a convex method to learn the 2-norm soft margin parameter in support vector machines, solving another important open problem. Finally, the novel approach presented in the paper is supported by positive empirical results.

Anomalous event detection using a semi-two dimensional hidden Markov model

The rapid increase in the deployment of CCTV systems has led to a greater demand for algorithms that are able to process incoming video feeds. These algorithms are designed to extract information of interest for human operators. During the past several years, there has been a large effort to detect abnormal activities through computer vision techniques. Typically, the problem is formulated as a novelty detection task where the system is trained on normal data and is required to detect events which do not fit the learned `normal' model. Many researchers have tried various sets of features to train different learning models to detect abnormal behaviour in video footage. In this work we propose using a Semi-2D Hidden Markov Model (HMM) to model the normal activities of people. The outliers of the model with insufficient likelihood are identified as abnormal activities. Our Semi-2D HMM is designed to model both the temporal and spatial causalities of the crowd behaviour by assuming the current state of the Hidden Markov Model depends not only on the previous state in the temporal direction, but also on the previous states of the adjacent spatial locations. Two different HMMs are trained to model both the vertical and horizontal spatial causal information. Location features, flow features and optical flow textures are used as the features for the model. The proposed approach is evaluated using the publicly available UCSD datasets and we demonstrate improved performance compared to other state of the art methods.

Successful outcomes in vocational education and training courses : how pedagogy and expectations influence achievement

This paper reports on a four year Australian Research Council funded Linkage Project titled Skilling Indigenous Queensland, conducted in regional areas of Queensland, Australia from 2009 to 2013. The project sought to investigate vocational education, training (VET) and teaching, Indigenous learners’ needs, employer cultural and expectations and community culture and expectations to identify best practice in numeracy teaching for Indigenous VET learners. Specifically it focused on ways to enhance the teaching and learning of courses and the associated mathematics in such courses to benefit learners and increase their future opportunities of employment. To date thirty-nine teachers/trainers/teacher aides and two hundred and thirty-one students consented to participate in the project. Nine VET courses were nominated to be the focus on the study. This paper focuses on questionnaire and interview responses from four trainers, two teacher aides and six students. In recent years a considerable amount of funding has been allocated to increasing Indigenous Peoples’ participation in education and employment. This increased funding is predicated on the assumption that it will make a difference and contribute to closing the education gap between Indigenous and non-Indigenous Australians (Council of Australia Governments, 2009). The central tenet is that access to education for Indigenous People will create substantial social and economic benefits for regional and remote Indigenous People. The project’s aim is to address some of the issues associated with the gap. To achieve the aims, the project adopted a mixed methods design aimed at benefitting research participants and included: participatory collaborative action research (Kemmis & McTaggart, 1988) and, community research (Smith, 1999). Participatory collaborative action research refers to a is a “collective, self-reflective enquiry undertaken by participants in social situations in order to improve the rationality and justice of their own social and educational practices” (Kemmis et al., 1988, p. 5). Community research is described as an approach that “conveys a much more intimate, human and self-defined space” (p. 127). Community research relies on and validates the community’s own definitions. As the project is informed by the social at a community level, it is described as “community action research or emancipatory research” (Smith, 1999, p. 127). It seeks to demonstrate benefit to the community, making positive differences in the lives of Indigenous People and communities. The data collection techniques included survey questionnaires, video recording of teaching and learning processes, teacher reflective video analysis of teaching, observations, semi-structured interviews and student numeracy testing. As a result of these processes, the findings indicate that VET course teachers work hard to adopt contextualising strategies to their teaching, however this process is not always straight forward because of the perceptions of how mathematics has been taught and learned historically. Further teachers, trainers and students have high expectations of one another with the view to successful outcomes from the courses.

Is foot ulcer simulation training (FUST) really effective? Participants’ supervisors speak out

Background Foot ulcers are a common reason for diabetes-related hospitalisation. Foot ulcer simulation training (FUST) programs have increased podiatry participants self-confidence to manage foot ulcers. However, supervisors’ perspectives on their participants attending these simulation programs have not been investigated. This mixed method (quantitative and qualitative) study aimed to investigate home clinical supervisors’ perspectives on any changes to their participants’ competence and practice following FUST. Methods Clinical supervisors of fifteen podiatrists, who participated in a two-day Foot Ulcer Simulation Training (FUST) course, were recruited. Supervisors completed quantitative surveys evaluating their participants’ foot ulcer competence pre-FUST and 6-months post-FUST, via a purposed designed 21-item survey using a five-point Likert scale (1=Very limited, 5=Highly competent). Supervisors also attended a semi-structured qualitative group interview to investigate supervisors’ perspectives on FUST. Results Supervisors surveys returned were pre-FUST (n=10) and post-FUST (n=12). Significant competence improvements were observed at the 6-month survey (mean scores 2.84 cf. 3.72, p < 0.05). Five supervisors attended the group interview. Five sub-themes emerged: i) FUST provided a good foundation for future learning, ii) FUST modelled good clinical behaviour, iii) clinical practice improvement was evident in most participants, iv) clinical improvements were dependent on participant’s willingness to change and existing workplace culture, v) FUST needs to be reinforced back in the home clinic. Conclusion Overall, supervisors of FUST participants indicated that the course improved their participants’ competence and clinical practice. However, the degree of improvement appears dependant on the participants’ home workplace culture and willingness to embrace change.

Development and evaluation of training in culturally specific screening and brief intervention for hospital patients with alcohol-related injuries

Objective To evaluate health practitioners’ confidence and knowledge of alcohol screening, brief intervention and referral after training in a culturally adapted intervention on alcohol misuse and well-being issues for trauma patients. Design Mixed methods, involving semi-structured interviews at baseline and a post-workshop questionnaire. Setting: Targeted acute care within a remote area major tertiary referral hospital. Participants Ten key informants and 69 questionnaire respondents from relevant community services and hospital-based health care professionals. Intervention Screening and brief intervention training workshops and resources for 59 hospital staff. Main outcome measures Self-reported staff knowledge of alcohol screening, brief intervention and referral, and satisfaction with workshop content and format. Results After training, 44% of participants reported being motivated to implement alcohol screening and intervention. Satisfaction with training was high, and most participants reported that their knowledge of screening and brief intervention was improved. Conclusion Targeted educational interventions can improve the knowledge and confidence of inpatient staff who manage patients at high risk of alcohol use disorder. Further research is needed to determine the duration of the effect and influence on practice behaviour. Ongoing integrated training, linked with systemic support and established quality improvement processes, is required to facilitate sustained change and widespread dissemination.

Understanding video based parent training intervention for children with autism

This paper investigates the challenges of delivering parent training intervention for autism over video. We conducted a qualitative field study of an intervention, which is based on a well-established training program for parents of children with autism, called Hanen More Than Words. The study was conducted with a Hanen Certified speech pathologist who delivered video based training to two mothers, each with a son having autism. We conducted observations of 14 sessions of the intervention spanning 3 months along with 3 semi-structured interviews with each participant. We identified different activities that participants performed across different sessions and analysed them based upon their implications on technology. We found that all the participants welcomed video based training but they also faced several difficulties, particularly in establishing rapport with other participants, inviting equal participation, and in observing and providing feedback on parent-child interactions. Finally, we reflect on our findings and motivate further investigations by defining three design sensitivities of Adaptation, Group Participation, and Physical Setup.

Semi-supervised learning of WLAN radio maps

In this thesis a manifold learning method is applied to the problem of WLAN positioning and automatic radio map creation. Due to the nature of WLAN signal strength measurements, a signal map created from raw measurements results in non-linear distance relations between measurement points. These signal strength vectors reside in a high-dimensioned coordinate system. With the help of the so called Isomap-algorithm the dimensionality of this map can be reduced, and thus more easily processed. By embedding position-labeled strategic key points, we can automatically adjust the mapping to match the surveyed environment. The environment is thus learned in a semi-supervised way; gathering training points and embedding them in a two-dimensional manifold gives us a rough mapping of the measured environment. After a calibration phase, where the labeled key points in the training data are used to associate coordinates in the manifold representation with geographical locations, we can perform positioning using the adjusted map. This can be achieved through a traditional supervised learning process, which in our case is a simple nearest neighbors matching of a sampled signal strength vector. We deployed this system in two locations in the Kumpula campus in Helsinki, Finland. Results indicate that positioning based on the learned radio map can achieve good accuracy, especially in hallways or other areas in the environment where the WLAN signal is constrained by obstacles such as walls.

Robust Semi-Blind Estimation for Beamforming Based MIMO Wireless Communication

In this paper, we present robust semi-blind (SB) algorithms for the estimation of beamforming vectors for multiple-input multiple-output wireless communication. The transmitted symbol block is assumed to comprise of a known sequence of training (pilot) symbols followed by information bearing blind (unknown) data symbols. Analytical expressions are derived for the robust SB estimators of the MIMO receive and transmit beamforming vectors. These robust SB estimators employ a preliminary estimate obtained from the pilot symbol sequence and leverage the second-order statistical information from the blind data symbols. We employ the theory of Lagrangian duality to derive the robust estimate of the receive beamforming vector by maximizing an inner product, while constraining the channel estimate to lie in a confidence sphere centered at the initial pilot estimate. Two different schemes are then proposed for computing the robust estimate of the MIMO transmit beamforming vector. Simulation results presented in the end illustrate the superior performance of the robust SB estimators.

Semi-supervised SVMs for classification with unknown class proportions and a small labeled dataset

In the design of practical web page classification systems one often encounters a situation in which the labeled training set is created by choosing some examples from each class; but, the class proportions in this set are not the same as those in the test distribution to which the classifier will be actually applied. The problem is made worse when the amount of training data is also small. In this paper we explore and adapt binary SVM methods that make use of unlabeled data from the test distribution, viz., Transductive SVMs (TSVMs) and expectation regularization/constraint (ER/EC) methods to deal with this situation. We empirically show that when the labeled training data is small, TSVM designed using the class ratio tuned by minimizing the loss on the labeled set yields the best performance; its performance is good even when the deviation between the class ratios of the labeled training set and the test set is quite large. When the labeled training data is sufficiently large, an unsupervised Gaussian mixture model can be used to get a very good estimate of the class ratio in the test set; also, when this estimate is used, both TSVM and EC/ER give their best possible performance, with TSVM coming out superior. The ideas in the paper can be easily extended to multi-class SVMs and MaxEnt models.

On the optimal usage of labelled examples in semi-supervised multi-class classification problems

In recent years, the performance of semi-supervised learning has been theoretically investigated. However, most of this theoretical development has focussed on binary classification problems. In this paper, we take it a step further by extending the work of Castelli and Cover [1] [2] to the multi-class paradigm. Particularly, we consider the key problem in semi-supervised learning of classifying an unseen instance x into one of K different classes, using a training dataset sampled from a mixture density distribution and composed of l labelled records and u unlabelled examples. Even under the assumption of identifiability of the mixture and having infinite unlabelled examples, labelled records are needed to determine the K decision regions. Therefore, in this paper, we first investigate the minimum number of labelled examples needed to accomplish that task. Then, we propose an optimal multi-class learning algorithm which is a generalisation of the optimal procedure proposed in the literature for binary problems. Finally, we make use of this generalisation to study the probability of error when the binary class constraint is relaxed.

Semi-supervised video segmentation using tree structured graphical models

We present a novel, implementation friendly and occlusion aware semi-supervised video segmentation algorithm using tree structured graphical models, which delivers pixel labels alongwith their uncertainty estimates. Our motivation to employ supervision is to tackle a task-specific segmentation problem where the semantic objects are pre-defined by the user. The video model we propose for this problem is based on a tree structured approximation of a patch based undirected mixture model, which includes a novel time-series and a soft label Random Forest classifier participating in a feedback mechanism. We demonstrate the efficacy of our model in cutting out foreground objects and multi-class segmentation problems in lengthy and complex road scene sequences. Our results have wide applicability, including harvesting labelled video data for training discriminative models, shape/pose/articulation learning and large scale statistical analysis to develop priors for video segmentation. © 2011 IEEE.