Multidimensional Learning Research Articles

Survey on Multi-Output Learning.

The aim of multi-output learning is to simultaneously predict multiple outputs given an input. It is an important learning problem for decision-making since making decisions in the real world often involves multiple complex factors and criteria. In recent times, an increasing number of research studies have focused on ways to predict multiple outputs at once. Such efforts have transpired in different forms according to the particular multi-output learning problem under study. Classic cases of multi-output learning include multi-label learning, multi-dimensional learning, multi-target regression, and others. From our survey of the topic, we were struck by a lack in studies that generalize the different forms of multi-output learning into a common framework. This article fills that gap with a comprehensive review and analysis of the multi-output learning paradigm. In particular, we characterize the four Vs of multi-output learning, i.e., volume, velocity, variety, and veracity, and the ways in which the four Vs both benefit and bring challenges to multi-output learning by taking inspiration from big data. We analyze the life cycle of output labeling, present the main mathematical definitions of multi-output learning, and examine the field's key challenges and corresponding solutions as found in the literature. Several model evaluation metrics and popular data repositories are also discussed. Last but not least, we highlight some emerging challenges with multi-output learning from the perspective of the four Vs as potential research directions worthy of further studies.

Creatıng Infographics Based on the Bridge21 Model for Team-based and Technology-mediated Learning

Aim/Purpose: The main aim of this study was modeling a collaborative process for knowledge visualization, via the creation of infographics. Background: As an effective method for visualizing complex information, creating infographics requires learners to generate and cultivate a deep knowledge of content and enables them to concisely visualize and share this knowledge. This study investigates creating infographics as a knowledge visualization process for collaborative learning situations by integrating the infographic design model into the team-based and technology-mediated Bridge21 learning model. Methodology: This study was carried out using an educational design perspective by conducting three main cycles comprised of three micro cycles: analysis and exploration; design and construction; evaluation and reflection. The process and the scaffolding were developed and enhanced from cycle to cycle based on both qualitative and quantitative methods by using the infographic design rubric and researcher observations acquired during implementation. Respectively, twenty-three, twenty-four, and twenty-four secondary school students participated in the infographic creation process cycles. Contribution: This research proposes an extensive step-by-step process model for creating infographics as a method of visualization for learning. It is particularly relevant for working with complex information, in that it enables collaborative knowledge construction and sharing of condensed knowledge. Findings: Creating infographics can be an effective method for collaborative learning situations by enabling knowledge construction, visualization and sharing. The Bridge21 activity model constituted the spine of the infographic creation process. The content generation, draft generation, and visual and digital design generation components of the infographic design model matched with the investigate, plan and create phases of the Bridge21 activity model respectively. Improvements on infographic design results from cycle to cycle suggest that the revisions on the process model succeeded in their aims. The rise in each category was found to be significant, but the advance in visual design generation was particularly large. Recommendations for Practitioners: The effectiveness of the creation process and the quality of the results can be boosted by using relevant activities based on learner prior knowledge and skills. While infographic creation can lead to a focus on visual elements, the importance of wording must be emphasized. Being a multidimensional process, groups need guidance to ensure effective collaboration. Recommendation for Researchers: The proposed collaborative infographic creation process could be structured and evaluated for online learning environments, which will improve interaction and achievement by enhancing collaborative knowledge creation. Impact on Society: In order to be knowledge constructors, innovative designers, creative communicators and global collaborators, learners need to be surrounded by adequate learning environments. The infographic creation process offers them a multidimensional learning situation. They must understand the problem, find an effective way to collect information, investigate their data, develop creative and innovative perspectives for visual design and be comfortable for using digital creation tools. Future Research: The infographic creation process could be investigated in terms of required learner prior knowledge and skills, and could be enhanced by developing pre-practices and scaffolding.

An Automatic Classification and Clustering Algorithm for Online Learning Goals Based on Cognitive Thinking

To improve the learning effect of online learning, an online learning target automatic classification and clustering analysis algorithm based on cognitive thinking was proposed. It was applied to a multi-dimensional learning community. A new form of virtual learning community concept was proposed. The design ideas of its multi-dimensional learning environment were elaborated. Ontology technology was used to collect student learning process data. A cognitive diagnostic model for assessing student learning status was generated. Finally, through the cluster analysis technology, the registered students in the curriculum center were automatically divided into different levels of community groups. The results showed that the proposed algorithm for automatic classification and clustering of online learning targets had a good application effect in the learning community. Therefore, this method has practical application value.

Chicken Raising in a Diverse Finnish Classroom: Multidimensional Sustainability Learning

Social change requires new educational planning and sustainable teaching methods. Shaping an environment of care with animals as a part of the daily school life may produce such a change. In this article, we present a transdisciplinary study with the aim of exploring whether raising chickens in a classroom could promote learning, especially sustainability learning, and how. The study employs an ethnographic approach and we have analyzed the data according to interaction analysis. We collected the data in a culturally-diverse Finnish primary school class during May 2018. The data comprise field notes, videos and photographs from indoor and outdoor school activities; interviews and discussions with teachers and students; and, texts and artifacts that were made by students. The results show that having chickens in the classroom not only improved the students’ learning of biology, but also enhanced many other activities. The chicken project became part of a complex learning culture that met several of the aims of the curriculum and in many ways reached beyond the aim of merely learning science. The project became a natural part of sustainability education and promoted the acquisition of knowledge and skills in relation to the ecological and social dimensions of sustainability.

MAB-based two-tier learning algorithms for joint channel and power allocation in stochastic underwater acoustic communication networks

This study investigates the problem of joint channel and power allocation in stochastic underwater acoustic communication networks. And the multiarmed bandit theory is employed to model this problem which includes unknown variables. This study presents two two-tier learning algorithms, which do not need any prior environment information. In the upper learning, the user plays the predicted best strategy and learns the actual played strategy. In the lower learning, “outdated virtual learning information,” which can be obtained by the information of actual played strategy, is learned. The two-tier actual–virtual learning enormously enriches the learning information and effectively improves the learning ability. And multidimensional learning method is presented to ease the difficulty caused by the coupling of joint strategy. With the evolution of learning time, the emphasis of learning transfers from the channel power sub-strategy dimension to the entire power strategy dimension. Due to the specific learning manner, the algorithms have high tolerance about delay and non-complete information. Simulation results show high performance and adaptability of the proposed learning algorithms.

Influence of learning strategy on response time during complex value-based learning and choice.

Measurements of response time (RT) have long been used to infer neural processes underlying various cognitive functions such as working memory, attention, and decision making. However, it is currently unknown if RT is also informative about various stages of value-based choice, particularly how reward values are constructed. To investigate these questions, we analyzed the pattern of RT during a set of multi-dimensional learning and decision-making tasks that can prompt subjects to adopt different learning strategies. In our experiments, subjects could use reward feedback to directly learn reward values associated with possible choice options (object-based learning). Alternatively, they could learn reward values of options’ features (e.g. color, shape) and combine these values to estimate reward values for individual options (feature-based learning). We found that RT was slower when the difference between subjects’ estimates of reward probabilities for the two alternative objects on a given trial was smaller. Moreover, RT was overall faster when the preceding trial was rewarded or when the previously selected object was present. These effects, however, were mediated by an interaction between these factors such that subjects were faster when the previously selected object was present rather than absent but only after unrewarded trials. Finally, RT reflected the learning strategy (i.e. object-based or feature-based approach) adopted by the subject on a trial-by-trial basis, indicating an overall faster construction of reward value and/or value comparison during object-based learning. Altogether, these results demonstrate that the pattern of RT can be informative about how reward values are learned and constructed during complex value-based learning and decision making.

Measuring what matters: Using technology to assess multidimensional learning

AbstractIn this article we describe the challenges associated with assessing complex competencies envisioned as the targets of learning in the 21st century. Such competencies typically reflect the integration of multiple dimensions of knowledge and skill. Technology plays a crucial role in their assessment, from conceptualisation through design, data gathering and interpretation of results. We use the case of science proficiency to illustrate challenges associated with the assessment of the intended products of multidimensional learning and the benefits provided by technology. We frame assessment development as an evidence‐centered design process and illustrate it by using cases drawn from middle school science. We then turn to ways in which assessment systems need to evolve to expand the scope of what can be done in the creation and use of valid, reliable and equitable assessments of complex, multidimensional learning. We conclude by discussing policy implications of technology‐based assessment systems with an emphasis on measuring what matters versus measuring what is easy, since what we choose to assess will become the focus of instruction. Major advances in assessment policy and practice require investment in the development, validation and deployment of technology‐based assessments that reflect the multidimensional competencies identified by contemporary research and theory.

Teaching Management System with Applications of RFID and IoT Technology

Currently, Internet of Things (IoT) technologies are used in many areas, such as intelligent transportation, smart city, hospital, games, education. Earlier interactive response system uses infrared or radio frequency (RF) wireless communication technologies to transmit the students’ answer to teachers’ managerment system, where there exists high cost, inconvenient usage, difficult deployment. How to use IoT to improve the quality of higher education becomes a very important topic in the researh area of teaching. Radio Frequency Identification (RFID) is one of key technologies to implement IoT applications, and most of universities use the High Frequency (HF) RFID card as the students’ identification devices in China. In this paper, a kind of WiFi supported RFID reader (WiRF) is implemented using open source hardware platforms, such as Node MCU and RFID-RC522. Then the proposed WiRF system is used to assist teacher to perform automatic attendance record and students’ behavior record. In addition, Quick Response (QR) code is another technology to enable IoT. In this paper, QR code is designed to quickly access course video and perform real-time interactive response in the classroom, which will provide multidimensional learning and strengthen the motivation of students’ learning. This IoT system can improve the attendance of students, and give a positive impact on students’ learning process for higher education.

Digital training platform for interpreting radiographic images of the chest

IntroductionTime delays and errors exist which lead to delays in patient care and misdiagnosis. Reporting clinicians follow guidance to form their own search strategy. However, little research has tested these training guides. With the use of eye tracking technology and expert input we developed a digital training platform to be used in chest image interpretation learning. MethodsTwo sections of a digital training platform were planned and developed; A) a search strategy training tool to assist reporters during their interpretation of images, and B) an educational tool to communicate the search strategies of expert viewers to trainees by using eye tracking technology. ResultsA digital training platform for use in chest image interpretation was created based on evidence within the literature, expert input and two search strategies previously used in clinical practice. Images and diagrams, aiding translation of the platform content, were incorporated where possible. The platform is structured to allow the chest image interpretation process to be clear, concise and methodical. ConclusionA search strategy was incorporated within the tool to investigate its use, with the possibility that it could be recommended as an evidence based approach for use by reporting clinicians. Eye tracking, a checklist and voice recordings have been combined to form a multi-dimensional learning tool, which has never been used in chest image interpretation learning before. The training platform for use in chest image interpretation learning has been designed, created and digitised. Future work will establish the efficacy of the developed approaches.

Metodologiczne aspekty stosowania techniki eye trackingowej w badaniach edukacyjnych

Educational studies focus on multidimensional learning and teaching processes and require the use of complex research methods for that reason. The development of new technologies and the increasing availability of equipment that previously was used in medical and biological science, enable extending a set of traditional social and humanities sciences research methods. These new technologies allow for the verification of the information obtained as a result of carrying out the diagnostic survey for example. One of these is eye tracking, a technique of tracking of eye movements that allows to infer about cognitive processes related to learning. This paper discusses the issue of application of the eye tracking technology in educational research. The article presents, inter alia, the most important research procedure assumptions, the qualitative and quantitative analysis methods of the eye movements measurement data, and the advantages and limitations of this technique.

Informative tools for characterizing individual differences in learning: Latent class, latent profile, and latent transition analysis

This article gives an introduction to latent class, latent profile, and latent transition models for researchers interested in investigating individual differences in learning and development. The models allow analyzing how the observed heterogeneity in a group (e.g., individual differences in conceptual knowledge) can be traced back to underlying homogeneous subgroups (e.g., learners differing systematically in their developmental phases). The estimated parameters include a characteristic response pattern for each subgroup, and, in the case of longitudinal data, the probabilities of transitioning from one subgroup to another over time. This article describes the steps involved in using the models, gives practical examples, and discusses limitations and extensions. Overall, the models help to characterize heterogeneous learner populations, multidimensional learning outcomes, non-linear learning pathways, and changing relations between learning processes. The application of these models can therefore make a substantial contribution to our understanding of learning and individual differences.

Density-based particle swarm optimization algorithm for data clustering

Abstract Particle swarm optimization (PSO) algorithm is widely used in cluster analysis. However, it is a stochastic technique that is vulnerable to premature convergence to sub-optimal clustering solutions. PSO-based clustering algorithms also require tuning of the learning coefficient values to find better solutions. The latter drawbacks can be evaded by setting a proper balance between the exploitation and exploration behaviors of particles while searching the feature space. Moreover, particles must take into account the magnitude of movement in each dimension and search for the optimal solution in the most populated regions in the feature space. This study presents a novel approach for data clustering based on particle swarms. In this proposal, the balance between exploitation and exploration processes is considered using a combination of (i) kernel density estimation technique associated with new bandwidth estimation method to address the premature convergence and (ii) estimated multidimensional gravitational learning coefficients. The proposed algorithm is compared with other state-of-the-art algorithms using 11 benchmark datasets from the UCI Machine Learning Repository in terms of classification accuracy, repeatability represented by the standard deviation of the classification accuracy over different runs, and cluster compactness represented by the average Dunn index values over different runs. The results of Friedman Aligned-Ranks test with Holm's test over the average classification accuracy and Dunn index values indicate that the proposed algorithm achieves better accuracy and compactness when compared with other algorithms. The significance of the proposed algorithm is represented in addressing the limitations of the PSO-based clustering algorithms to push forward clustering as an important technique in the field of expert systems and machine learning. Such application, in turn, enhances the classification accuracy and cluster compactness. In this context, the proposed algorithm achieves better results compared with other state-of-the-art algorithms when applied to high-dimensional datasets (e.g., Landsat and Dermatology). This finding confirms the importance of estimating multidimensional learning coefficients that consider particle movements in all the dimensions of the feature space. The proposed algorithm can likewise be applied in repeatability matters for better decision making, as in medical diagnosis, as proved by the low standard deviation obtained using the proposed algorithm in conducted experiments.

Structured Constructs Models Based on Change-Point Analysis

This study proposes a structured constructs model (SCM) to examine measurement in the context of a multidimensional learning progression (LP). The LP is assumed to have features that go beyond a typical multidimentional IRT model, in that there are hypothesized to be certain cross-dimensional linkages that correspond to requirements between the levels of the different dimensions. The new model builds on multidimensional item response theory models and change-point analysis to add cut-score and discontinuity parameters that embody these substantive requirements. This modeling strategy allows us to place the examinees in the appropriate LP level and simultaneously to model the hypothesized requirement relations. Results from a simulation study indicate that the proposed change-point SCM recovers the generating parameters well. When the hypothesized requirement relations are ignored, the model fit tends to become worse, and the model parameters appear to be more biased. Moreover, the proposed model can be used to find validity evidence to support or disprove initial theoretical hypothesized links in the LP through empirical data. We illustrate the technique with data from an assessment system designed to measure student progress in a middle-school statistics and modeling curriculum.

A new comprehensive model for Continuous Professional Development

Background: It is generally agreed that continuing professional development (CPD) for GPs is important for quality of care. Internationally, however, different approaches to identify the learning objectives and the CPD content have been chosen.Objectives: To improve GPs’ CPD in Denmark we explore how general practitioners’ (GPs) self-experienced learning needs can be combined with learning needs experienced from a societal perspective and still make sense for GPs.Methods: We performed a multi-dimensional learning needs analysis with a modified Delphi method in a participatory action research set-up. Twenty practice-based small learning groups and a group appointed by the Danish public health service were asked to identify learning needs with the Danish family medicine curriculum as reference. Then we asked a group of GP researchers and hospital consultants, a group of GPs with interests in narrative, person-centred medicine and a group of GP educators, and administrative staff, to triangulate the initial findings.Results: We identified educational themes through a defined collaborative consensus oriented process. Examples of themes are the diagnostic challenge, care for patients with multi-morbidity, elderly patients and children. Due to variation in requested learning objectives, the identified themes do not cover all relevant areas for CPD training. The identified themes will only make sense if seen as supplementary to other CPD activities based on GPs individual needs analyses.Conclusion: It is possible to identify prioritized educational themes for GPs through a process involving the majority of stakeholders. Nevertheless, CPD should also include activities based on individual needs analysis.KEY MESSAGES When developing a continuing professional development programme for GPs, educational themes can be identified by a multi-dimensional needs analysis involving different stakeholders.GPs might benefit from activities based on a consensus-based curriculum combined with an individual needs analysis, thus exposing individual GPs to both recognized and non-recognized relevant learning needs.

Toward Distributed Data Processing on Intelligent Leak-Points Prediction in Petrochemical Industries

Focusing on the leak-points in petrochemical industries, this paper discusses the key factors (i.e., equipment temperature, gas pressure, and diffusion rate) in petrochemical industries. Data from sensors of petrochemical industries need to be timely operated because of time sensitivity and it is hard to achieve associated information from sensors located in production sites. To this end, we propose a three-level framework based on improved back propagation (TLBP). The real-time data streams are processed according to the arriving time in input layer. At the same time, a neuron-optimizing solution is introduced in learning process to deal with redundant and invalid neurons, thereby accelerating the response speed of learning and reducing the prediction time. Finally, we propose an improved mechanism of the multidimensional learning factor to lower the learning error and higher convergence rate. Meanwhile, to fulfill the distributed prediction on leak-points, we see one three-level data-processing unit as a logic machine with multiple operators. Using the assignment scheduling, the general scheduling problem is split into the common subproblem of every operator and the system overhead is reduced. With the processed data we can obtain the relative location or diffusion radius of leak-points, as well as the area of leak-points. Simulation results show that the TLBP performs better than related algorithms in different metrics. Besides, the adaptability of TLBP is verified in leak-points prediction of petrochemical equipment from the processed data.

Learner Attrition in an Advanced Vocational Online Training: The Role of Computer Attitude, Computer Anxiety, and Online Learning Experience

Abstract Online learning has gained importance in education over the last 20 years, but the well-known problem of high dropout rates still persists. According to the multi-dimensional learning tasks model, the cognitive (over)load of learners is essential to attrition when dealing with five challenges (e.g. technology, user interface) of an online training (Tyler-Smith, 2006). The experienced load might depend on learner characteristics. The study explored the extent that learners dropping out from a vocational video-based online training about media design for employees of micro, small and medium-sized enterprises differ from working learners’ online learning experience, computer attitude, and computer anxiety. The data were collected from 72 of 128 registered employees who completed a questionnaire before starting the course to analyze differences between the dropout group (submitted no solutions to online training tasks; n = 19) and the active learner group (submitted at least one of 13 task solutions; n = 53). No differences were found in online learning experience, but the dropout group reported more negative attitudes towards computers and a higher level of computer anxiety than the active learner group.

Authentic learning for teaching reading: Foundation phase pre-service student teachers’ learning experiences of creating and using digital stories in real classrooms

Teaching and learning, an evolving endeavour, is associated with many factors, with advancements in technology, playing an ever-growing role in the classroom. It is therefore important to include the use of interactive communication technologies (ICTs) in university curricula of teacher education programmes. Universities ought to be creative in advancing autonomous learning among their students by providing opportunities for integrated and rich learning experiences. Accordingly, the purpose of this study was to intentionally integrate ICTs in the planning and delivery of foundation phase reading lessons. This was achieved by providing authentic learning opportunities to final year foundation phase student teachers through the provision of training in the creation of digital stories (DS), collaborating within communities of practice (COP) (peers and other relevant parties), and then using their creations in ‘real-world’ classroom contexts. The aims of this study were to explore student teachers’ perceptions and experiences of developing DS in groups with minimal formal initial input and their use of DS during foundation phase (FP) reading lessons in real-class settings during teaching practice. Data were collected via focus group interviews and participants’ reflection essays. The study’s findings indicate that the creation of their own DS provided rich, rewarding multidimensional learning experiences to student teachers. Participants reported that they found the ‘assignment’ to be of real value, since it was directly linked to classroom practice, and despite the cognitive demands of the assignment; the nature of the task nurtured, an agentic disposition towards their own learning. Participants further reported that the DS provided enthusiasm among young learners during the delivery of lessons and were of pedagogical value, despite experiencing some challenges in using DS during reading lessons. Participants were of the view that the use of DS in advancing reading and literacy holds much pedagogical promise, because it resonates with the this generation of digital natives, the present generation of learners who have been born into a world where they interact with digital technology from an early age.

Necessary Contributions of Human Frontal Lobe Subregions to Reward Learning in a Dynamic, Multidimensional Environment.

Real-world decisions are typically made between options that vary along multiple dimensions, requiring prioritization of the important dimensions to support optimal choice. Learning in this setting depends on attributing decision outcomes to the dimensions with predictive relevance rather than to dimensions that are irrelevant and nonpredictive. This attribution problem is computationally challenging, and likely requires an interplay between selective attention and reward learning. Both these processes have been separately linked to the prefrontal cortex, but little is known about how they combine to support learning the reward value of multidimensional stimuli. Here, we examined the necessary contributions of frontal lobe subregions in attributing feedback to relevant and irrelevant dimensions on a trial-by-trial basis in humans. Patients with focal frontal lobe damage completed a demanding reward learning task where options varied on three dimensions, only one of which predicted reward. Participants with left lateral frontal lobe damage attributed rewards to irrelevant dimensions, rather than the relevant dimension. Damage to the ventromedial frontal lobe also impaired learning about the relevant dimension, but did not increase reward attribution to irrelevant dimensions. The results argue for distinct roles for these two regions in learning the value of multidimensional decision options under dynamic conditions, with the lateral frontal lobe required for selecting the relevant dimension to associate with reward, and the ventromedial frontal lobe required to learn the reward association itself. The real world is complex and multidimensional; how do we attribute rewards to predictive features when surrounded by competing cues? Here, we tested the critical involvement of human frontal lobe subregions in a probabilistic, multidimensional learning environment, asking whether focal lesions affected trial-by-trial attribution of feedback to relevant and irrelevant dimensions. The left lateral frontal lobe was required for filtering option dimensions to allow appropriate feedback attribution, while the ventromedial frontal lobe was necessary for learning the value of features in the relevant dimension. These findings argue that selective attention and associative learning processes mediated by anatomically distinct frontal lobe subregions are both critical for adaptive choice in more complex, ecologically valid settings.

Multiscale Support Vector Learning With Projection Operator Wavelet Kernel for Nonlinear Dynamical System Identification.

A giant leap has been made in the past couple of decades with the introduction of kernel-based learning as a mainstay for designing effective nonlinear computational learning algorithms. In view of the geometric interpretation of conditional expectation and the ubiquity of multiscale characteristics in highly complex nonlinear dynamic systems [1]-[3], this paper presents a new orthogonal projection operator wavelet kernel, aiming at developing an efficient computational learning approach for nonlinear dynamical system identification. In the framework of multiresolution analysis, the proposed projection operator wavelet kernel can fulfill the multiscale, multidimensional learning to estimate complex dependencies. The special advantage of the projection operator wavelet kernel developed in this paper lies in the fact that it has a closed-form expression, which greatly facilitates its application in kernel learning. To the best of our knowledge, it is the first closed-form orthogonal projection wavelet kernel reported in the literature. It provides a link between grid-based wavelets and mesh-free kernel-based methods. Simulation studies for identifying the parallel models of two benchmark nonlinear dynamical systems confirm its superiority in model accuracy and sparsity.

Rational Polarization: Asymptotic Belief Divergence in Response to Common Information

We provide a model of boundedly rational, multidimensional learning and characterize when beliefs will converge to the truth. Agents maintain beliefs as marginal probabilities instead of joint probabilities, and agents' information is of lower dimension than the model. As a result, for some observations, agents may face an identification problem affecting the role of data in inference. Beliefs converge to the truth when these observations are rare, but beliefs diverge when observations presenting an identification problem are frequent. Robustly, two agents with differing priors who observe identical, unambiguous information may disagree forever, with stronger disagreement the more information received.