Incremental learning from stream data

Recent years have witnessed an incredibly increasing interest in the topic of incremental learning. Unlike conventional machine learning situations, data flow targeted by incremental learning becomes available continuously over time. Accordingly, it is desirable to be able to abandon the traditional assumption of the availability of representative training data during the training period to develop decision boundaries. Under scenarios of continuous data flow, the challenge is how to transform the vast amount of stream raw data into information and knowledge representation, and accumulate experience over time to support future decision-making process. In this paper, we propose a general adaptive incremental learning framework named ADAIN that is capable of learning from continuous raw data, accumulating experience over time, and using such knowledge to improve future learning and prediction performance. Detailed system level architecture and design strategies are presented in this paper. Simulation results over several real-world data sets are used to validate the effectiveness of this method.

The Critical Reading of the Images Associated with Science-Related News Reports: Establishing a knowledge, skills, and attitudes framework

For the majority of adults, the media constitute their main source of information about science and science-related matters impacting on society. To help prepare young people to engage with science in the media, teachers are being exhorted to equip their students with the knowledge, skills, and attitudes to respond critically to science-related news reports. Typically, such reports comprise not only text, but also visual elements. These images are not simply adjuncts to the written word; they are integral to meaning-making. Though science teachers make considerable use of newspaper images, they tend to view these representations unproblematically, underestimating their potential ambiguity, complexity, and role in framing media messages. They rarely aim to develop students’ ability to ‘read’, critically, such graphics. Moreover, research into how this might be achieved is limited and, consequently, research-informed guidance which could support this instruction is lacking. This paper describes a study designed to formulate a framework for such teaching. Science communication scholars, science journalists and media educators with acknowledged relevant expertise were surveyed to ascertain what knowledge, skills, and attitudes they deemed useful to engagement with science related news images. Their proposals were recast as learning intentions (instructional objectives), and science and English teachers collaborated to suggest which could be addressed with secondary school students and the age group best suited to their introduction. The outcome is an inventory of learning intentions on which teachers could draw to support their planning of instructional sequences aimed at developing students’ criticality in respect of the totality of science news reports.

Robust visual tracking based on online learning sparse representation

Handling appearance variations is a very challenging problem for visual tracking. Existing methods usually solve this problem by relying on an effective appearance model with two features: (1) being capable of discriminating the tracked target from its background, (2) being robust to the target's appearance variations during tracking. Instead of integrating the two requirements into the appearance model, in this paper, we propose a tracking method that deals with these problems separately based on sparse representation in a particle filter framework. Each target candidate defined by a particle is linearly represented by the target and background templates with an additive representation error. Discriminating the target from its background is achieved by activating the target templates or the background templates in the linear system in a competitive manner. The target's appearance variations are directly modeled as the representation error. An online algorithm is used to learn the basis functions that sparsely span the representation error. The linear system is solved via ℓ1 minimization. The candidate with the smallest reconstruction error using the target templates is selected as the tracking result. We test the proposed approach using four sequences with heavy occlusions, large pose variations, drastic illumination changes and low foreground-background contrast. The proposed approach shows excellent performance in comparison with two latest state-of-the-art trackers.

Live Project love: building a framework for Live Projects

Live projects adopt a wide range of approaches: design/ build, community engagement, participation, protest, analysis, etc. They are driven by tutors with passion, expertise and the ability to manage risk, in ways that exhibit fluency and high levels of skill. They also offer sites of student-led and community co-learning, can support research, demonstrate ‘impact’ and satisfy universities’ policies on outreach. Whilst the breadth and reach of Live Projects is now demonstrably wide, we still fail to fully locate Live Projects within a pedagogical context, tending instead to limit our descriptions and hence analysis to the architectural process and outcome. This failure to locate Live Projects within a pedagogical framework means we still struggle to encapsulate, critique, progress, and indeed, elevate the work.

This chapter draws on some of the case studies presented in recent papers and international conferences in order to provide educators with signposts and important overviews around which and in respect to they can develop their own pedagogical frameworks.

Semi batch learning with store management using enhanced conjugate gradient

This paper explores the performance of sliding-window based training, termed as semi batch, using multilayer perceptron (MLP) neural network in the presence of correlated data. The sliding window training is a form of higher order instantaneous learning strategy without the need of covariance matrix, usually employed for modeling and tracking purposes. Sliding-window framework is implemented to combine the robustness of offline learning algorithms with the ability to track online the underlying process of a function. This paper adopted sliding window training with recent advances in conjugate gradient direction with application of data store management e.g. simple distance measure, angle evaluation and the novel prediction error test. The simulation results show the best convergence performance is gained by using store management techniques. © 2012 Springer-Verlag.

Sonification as Concurrent Augmented Feedback for Motor Skill Learning and the Importance of Mapping Design

In recent years, sonification of movement has emerged as a viable method for the provision of feedback in motor learning. Despite some experimental validation of its utility, controlled trials to test the usefulness of sonification in a motor learning context are still rare. As such, there are no accepted conventions for dealing with its implementation. This article addresses the question of how continuous movement information should be best presented as sound to be fed back to the learner. It is proposed that to establish effective approaches to using sonification in this context, consideration must be given to the processes that underlie motor learning, in particular the nature of the perceptual information available to the learner for performing the task at hand. Although sonification has much potential in movement performance enhancement, this potential is largely unrealised as of yet, in part due to the lack of a clear framework for sonification mapping: the relationship between movement and sound. By grounding mapping decisions in a firmer understanding of how perceptual information guides learning, and an embodied cognition stance in general, it is hoped that greater advances in use of sonification to enhance motor learning can be achieved.

Local Conventions in Game Play in an Evolving Dual Social Network Framework

People usually perform economic interactions within the social setting of a small group, while they obtain relevant information from a broader source. We capture this feature with a dynamic interaction model based on two separate social networks. Individuals play a coordination game in an interaction network, while updating their strategies using information from a separate influence network through which information is disseminated. In each time period, the interaction and influence networks co-evolve, and the individuals’ strategies are updated through a modified naive learning process. We show that both network structures and players’ strategies always reach a steady state, in which players form fully connected groups and converge to local conventions. We also analyze the influence exerted by a minority group of strongly opinionated players on these outcomes.

Effective Teaching and Learning of Science in the Outdoors: A Blended Model

This paper reports on an innovative Continuing Professional Development (CPD) programme which addressed transition issues and issues with conducting outdoor work and attitudes towards science through ‘Shared Learning' days between elementary and middle school transition classes. Teachers supported each other to overcome issues with conducting outdoor work and contributed their expertise from their educational stage. The project utilised a blended CPD approach of workshops, coteaching and in-class support and was based upon a wealth earlier successful CPD programmes to result in a sound theoretical framework.
The outcomes were measured using a thorough mixed-methods approach. This paper will report on the achieved outcomes with effective outdoor learning as the vehicle to overcome identified issues and key challenges for policy development.

Teaching Programming - Understanding Lecture Capture YouTube Analytics

ABSTRACT
The proliferation in the use of video lecture capture in universities worldwide presents an opportunity to analyse video watching patterns in an attempt to quantify and qualify how students engage and learn with the videos. It also presents an opportunity to investigate if there are similar student learning patterns during the equivalent physical lecture. The goal of this action based research project was to capture and quantitatively analyse the viewing behaviours and patterns of a series of video lecture captures across several university Java programming modules. It sought to study if a quantitative analysis of viewing behaviours of Lecture Capture videos coupled with a qualitative evaluation from the students and lecturers could be correlated to provide generalised patterns that could then be used to understand the learning experience of students during videos and potentially face to face lectures and, thereby, present opportunities to reflectively enhance lecturer performance and the students’ overall learning experience. The report establishes a baseline understanding of the analytics of videos of several commonly used pedagogical teaching methods used in the delivery of programming courses. It reflects on possible concurrences within live lecture delivery with the potential to inform and improve lecturing performance.

An Automated Design Framework for Assembly Task Simulation

This paper presents an automated design framework for the development of individual part forming tools for a composite stiffener. The framework uses parametrically developed design geometries for both the part and its layup tool. The framework has been developed with a functioning user interface where part / tool combinations are passed to a virtual environment for utility based assessment of their features and assemblability characteristics. The work demonstrates clear benefits in process design methods with conventional design timelines reduced from hours and days to minutes and seconds. The methods developed here were able to produce a digital mock up of a component with its associated layup tool in less than 3 minutes. The virtual environment presenting the design to the designer for interactive assembly planning was generated in 20 seconds. Challenges still exist in determining the level of reality required to provide an effective learning environment in the virtual world. Full representation of physical phenomena such as gravity, part clashes and the representation of standard build functions require further work to represent real physical phenomena more accurately.

Machine learning classification of surgical pathology reports and chunk recognition for information extraction noise reduction

Background and aims: Machine learning techniques for the text mining of cancer-related clinical documents have not been sufficiently explored. Here some techniques are presented for the pre-processing of free-text breast cancer pathology reports, with the aim of facilitating the extraction of information relevant to cancer staging.

Materials and methods: The first technique was implemented using the freely available software RapidMiner to classify the reports according to their general layout: ‘semi-structured’ and ‘unstructured’. The second technique was developed using the open source language engineering framework GATE and aimed at the prediction of chunks of the report text containing information pertaining to the cancer morphology, the tumour size, its hormone receptor status and the number of positive nodes. The classifiers were trained and tested respectively on sets of 635 and 163 manually classified or annotated reports, from the Northern Ireland Cancer Registry.

Results: The best result of 99.4% accuracy – which included only one semi-structured report predicted as unstructured – was produced by the layout classifier with the k nearest algorithm, using the binary term occurrence word vector type with stopword filter and pruning. For chunk recognition, the best results were found using the PAUM algorithm with the same parameters for all cases, except for the prediction of chunks containing cancer morphology. For semi-structured reports the performance ranged from 0.97 to 0.94 and from 0.92 to 0.83 in precision and recall, while for unstructured reports performance ranged from 0.91 to 0.64 and from 0.68 to 0.41 in precision and recall. Poor results were found when the classifier was trained on semi-structured reports but tested on unstructured.

Conclusions: These results show that it is possible and beneficial to predict the layout of reports and that the accuracy of prediction of which segments of a report may contain certain information is sensitive to the report layout and the type of information sought.

Towards a Resilience Metric Framework for Cyber-Physical Systems

Resilience is widely accepted as a desirable system property for cyber-physical systems. However, there are no metrics that can be used to measure the resilience of cyber-physical systems (CPS) while the multi-dimensional nature of performance in these systems is considered. In this work, we present first results towards a resilience metric framework. The key contributions of this framework are threefold: First, it allows to evaluate resilience with respect to different performance indicators that are of interest. Second, complexities that are relevant to the performance indicators of interest, can be intentionally abstracted. Third and final, it supports the identification of reasons for good or bad resilience to improve system design.

HRM and Innovation: A Multi-level Organizational Learning Perspective

Drawing on the 4I organizational learning framework (Crossan et al., 1999), this article develops a model to explain the multi-level and cross-level relationships between HRM practices and innovation. Individual, team, and organizational level learning stocks are theorized to explain how HRM practices affect innovation at a given level. Feed-forward and feedback learning flows explain how cross-level effects of HRM practices on innovation take place. In addition, we propose that HRM practices fostering individual, team, and organizational level learning should form a coherent system to facilitate the emergence of innovation. The article is concluded with discussions on its contributions and potential future research directions.