Active Learning with Statistical Models

For many types of learners one can compute the statistically 'optimal' way to select data. We review how these techniques have been used with feedforward neural networks. We then show how the same principles may be used to select data for two alternative, statistically-based learning architectures: mixtures of Gaussians and locally weighted regression. While the techniques for neural networks are expensive and approximate, the techniques for mixtures of Gaussians and locally weighted regression are both efficient and accurate.

Sequential Optimal Recovery: A Paradigm for Active Learning

In most classical frameworks for learning from examples, it is assumed that examples are randomly drawn and presented to the learner. In this paper, we consider the possibility of a more active learner who is allowed to choose his/her own examples. Our investigations are carried out in a function approximation setting. In particular, using arguments from optimal recovery (Micchelli and Rivlin, 1976), we develop an adaptive sampling strategy (equivalent to adaptive approximation) for arbitrary approximation schemes. We provide a general formulation of the problem and show how it can be regarded as sequential optimal recovery. We demonstrate the application of this general formulation to two special cases of functions on the real line 1) monotonically increasing functions and 2) functions with bounded derivative. An extensive investigation of the sample complexity of approximating these functions is conducted yielding both theoretical and empirical results on test functions. Our theoretical results (stated insPAC-style), along with the simulations demonstrate the superiority of our active scheme over both passive learning as well as classical optimal recovery. The analysis of active function approximation is conducted in a worst-case setting, in contrast with other Bayesian paradigms obtained from optimal design (Mackay, 1992).

A Formulation for Active Learning with Applications to Object Detection

We discuss a formulation for active example selection for function learning problems. This formulation is obtained by adapting Fedorov's optimal experiment design to the learning problem. We specifically show how to analytically derive example selection algorithms for certain well defined function classes. We then explore the behavior and sample complexity of such active learning algorithms. Finally, we view object detection as a special case of function learning and show how our formulation reduces to a useful heuristic to choose examples to reduce the generalization error.

Literature in the foreign language syllabus : engaging the student through active learning. 'El rol de la literatura en la enseñanza de la lengua extranjera : la implicación del estudiante en el aula a través del aprendizaje activo'.

Resumen en español. Resumen tomado de la publicación

Multidisciplinary analysis of clinical cases: a transition towards active learning in dental education

In the past 2009/10 academic year, we took steps towards introduction of active methodologies, from a multidisciplinar approach, into a conventional lecture-based Dental Education program. We consolidated these practices in the current 2010/11 year, already within a new Bologna-adapted scheme. Transition involved (i) critical assessment of the limitations of traditional teaching (ii) identification of specific learning topics allowing for integration of contents, (iii) implementation of student-centred learning activities in old curricular plans (iv) assessment of students' satisfaction and perceived learning outcomes, (v) implementation of these changes in new Bologna-adapted curricula

From activity to active learning: drawing and dialogue interaction in virtual learning environments for computer-supported collaborative work in design.

Virtual learning environments (VLEs) would appear to be particular effective in computer-supported collaborative work (CSCW) for active learning. Most research studies looking at computer-supported collaborative design have focused on either synchronous or asynchronous modes of communication, but near-synchronous working has received relatively little attention. Yet it could be argued that near-synchronous communication encourages creative, rhetorical and critical exchanges of ideas, building on each other’s contributions. Furthermore, although many researchers have carried out studies on collaborative design protocol, argumentation and constructive interaction, little is known about the interaction between drawing and dialogue in near-synchronous collaborative design. The paper reports the first stage of an investigation into the requirements for the design and development of interactive systems to support the learning of collaborative design activities. The aim of the study is to understand the collaborative design processes while sketching in a shared white board and audio conferencing media. Empirical data on design processes have been obtained from observation of seven sessions with groups of design students solving an interior space-planning problem of a lounge-diner in a virtual learning environment, Lyceum, an in-house software developed by the Open University to support its students in collaborative learning.

Semantic organization to enhance active learning based on distributed multimedia educational resources

In order to organize distributed educational resources efficiently, to provide active learners an integrated, extendible and cohesive interface to share the dynamically growing multimedia learning materials on the Internet, this paper proposes a generic resource organization model with semantic structures to improve expressiveness, scalability and cohesiveness. We developed an active learning system with semantic support for learners to access and navigate through efficient and flexible manner. We learning resources in an efficient and flexible manner. We provide facilities for instructors to manipulate the structured educational resources via a convenient visual interface. We also developed a resource discovering and gathering engine based on complex semantic associations for several specific topics.

Implementation of student active learning in primary mathematics in Indonesia

Twenty years after the first pilot projects began to develop Student Active Learning (SAL) in Indonesia, and four years since it was adopted for use in the last provinces, this research investigates the implementation of Student Active Learning in Indonesian primary mathematics classrooms. A study of the relevant literature indicates that teaching based on constructivist principles is unlikely to be implemented well in mathematics classrooms unless there are high quality teachers, readily available manipulative materials, and a supportive learning environment. As Indonesian schools often lack one or more of these aspects, it seemed likely that Student Active Learning principles might not be ‘fully’ implemented in Indonesian primary mathematics classrooms. Thus a smaller scale, parallel study was carried out in Australian schools where there is no policy of Student Active Learning, but where its underlying principles are compatible with the stated views about learning and teaching mathematics. The study employed a qualitative interpretive methodology. Sixteen primary teachers from four urban and four rural Indonesian schools and four teachers from two Victorian schools were observed for four mathematics lessons each. The twenty teachers, as well as fourteen Indonesian headteachers and other education professionals, were interviewed in order to establish links between the background and beliefs of participants, and their implementation of Student Active Learning. Information on perceived constraints on the implementation of SAL was also sought. The results of this study suggest that Student Active learning has been implemented at four levels in Indonesian primary mathematics classrooms, ranging from essentially no implementation to a relatively high level of implementation, with an even higher level of implementation in three of the four Australian classrooms observed. Indonesian teachers, headteachers and supervisors hold a range of views of SAL and also of mathematics learning and teaching. These views largely depended on their in-service training in SAL and, more particularly, on their participation in the PEQIP project Typically, participants’ expressed views of SAL were at the same or higher level as their views of mathematics learning and teaching, with a similar pattern being observed in the relationship between these latter views and their implementation of SAL principles. Three factors were identified as influencing teacher change in terms of implementation of SAL: policy, curricular and organisational, and attitudes. Recommendations arising from this study include the adoption of reflection as an underlying principle in the theory of SAL, the continuation and extension of PEQIP type projects, changes in government policy on curriculum coverage and pre-service teacher training, and more support for teachers at the school and local authority levels.

The development of teaching skills to support active learning in university science (ALIUS)

This paper describes an Australian Learning and Teaching Council funded project for which Learning Design is encompassed in the broadest sense. ALIUS (Active Learning In University Science) takes the design of learning back to the learning experiences created for students. ALIUS is not about designing a particular activity, or subject, or course, but rather the development of a method, or process, by which we have re-designed the way in which learning occurs in large university classrooms world wide.

Spamcooling : a parallel heterogeneous ensemble spam filtering system based on active learning techniques

Anti-spam technology is developing rapidly in recent years. With the emerging applications of machine learning in diverse fields, researchers as well as manufacturers around the world have attempted a large number of related algorithms to prevent spam. In this paper, we designed an effective anti-spam protection system, SpamCooling, based on the mechanism of active learning and parallel heterogeneous ensemble learning techniques. The system adopts a batch method to filter spam and can be easily incorporated with existing mail clients (MUA). It can actively obtain user feedbacks for providing users with personalized spam filtering experiences. The parallel heterogeneous ensemble method can help system achieve high spam detection rate as well as low ham misclassification rate.

Using e Technologies for active learning

This paper describes a study of educators at an Australian University who are using eTechnolo-gies in their teaching. The perceptions of the educators of the technologies are discussed as well as how they believe these technologies help students to learn. Outcomes from the study offer suggestions of how appropriate eTechnologies can be incorporated into online learning environments for effective teaching. Particular emphasis is made of eTechnologies which support active learning by students, one of the goals of the principles of good teaching practice in undergraduate education.

Philosophy has consequences! Developing metacognition and active learning in the ethics classroom

The importance of enchancing metacognition and encouraging active learning in philosophy teaching has been increasingly recognised in recent years. Yet traditional teaching methods have not always centralised helping students to become reflectively and critically aware of the quality and consistency of their own thinking. This is particularly relevant when teaching moral philosophy, where apparently inconsistent intuitions and responses are common. In this paper I discuss the theoretical basis of the relevance of metacognition and active learning for teaching moral philosophy. Applying recent discussions of metacognition, intuition conflicts and survey-based teaching techniques, I then outline a strategy for encouraging metacognitive awareness of tensions in students’ pretheoretical beliefs, and developing a critical self-awareness of their development as moral thinkers.

Active learning: the importance of developing a comprehensive measure

This article reports on an investigation into the validity of a widely used scale for measuring the extent to which higher education students employ active learning strategies. The scale is the active learning scale in the Australasian Survey of Student Engagement. This scale is based on the Active and Collaborative Learning scale of the National Survey of Student Engagement. The particular focus of the study was to investigate effects resulting from the addition of a small number of items to the active learning scale designed to capture some highly engaging, mostly online, activities. The items were developed in response to concerns that students studying in distance mode often report lower average scores on active learning scales than their on-campus counterparts. The additional items relate to activities such as working online with other students and faculty. The findings show that average scores on the Australasian Survey of Student Engagement/National Survey of Student Engagement scale increase significantly when the new items are included and that some differences between on-campus and distance education students narrow significantly. These findings have implications for the development of more robust and comprehensive instruments to measure active learning.