Search

Computing accurate yet efficient approximations to the solutions of the electronic Schrödinger equation has been a paramount challenge of computational chemistry for decades. Quantum Monte Carlo methods are a promising avenue of development as their core algorithm exhibits a number of favorable properties: it is highly parallel and scales favorably with the considered system size, with an accuracy that is limited only by the choice of the wave function Ansatz. The recently introduced machine-learned parametrizations of quantum Monte Carlo Ansätze rely on the efficiency of neural networks as universal function approximators to achieve state of the art accuracy on a variety of molecular systems. With interest in the field growing rapidly, there is a clear need for easy to use, modular, and extendable software libraries facilitating the development and adoption of this new class of methods. In this contribution, the DeepQMC program package is introduced, in an attempt to provide a common framework for future investigations by unifying many of the currently available deep-learning quantum Monte Carlo architectures. Furthermore, the manuscript provides a brief introduction to the methodology of variational quantum Monte Carlo in real space, highlights some technical challenges of optimizing neural network wave functions, and presents example black-box applications of the program package. We thereby intend to make this novel field accessible to a broader class of practitioners from both the quantum chemistry and the machine learning communities. [ABSTRACT FROM AUTHOR]

The Mathematical Sciences Department at Shenandoah University focused on increasing the participation of women in the mathematics program by fostering a welcoming learning community and promoting double majors. Data from the past decade is presented. During this time, the program saw an increase in women undergraduate mathematics majors from 11% to as high as 71%. This paper describes the initiatives taken to make these changes. [ABSTRACT FROM AUTHOR]

Clustering, a well-studied problem in the machine learning community, becomes even more intriguing with the emergence of quantum machine learning. Specifically, exploring clustering techniques for quantum data, such as quantum states, holds great interest. This paper introduces a quantum state clustering algorithm that utilizes variational quantum circuits. Our algorithm transforms the clustering problem into a parameter optimization task involving parametric quantum circuits. Each cluster is represented by a variational quantum circuit (VQC), which learns to extract the distinctive feature of its corresponding cluster during the optimization process. To guide the optimization of circuit parameters, we design an objective function that encourages each cluster's feature extractor to produce features similar to states within its own cluster and dissimilar to states in other clusters. We construct four quantum state datasets for testing the effectiveness of our algorithm. The numerical results demonstrate that our algorithm can achieve satisfying performance. [ABSTRACT FROM AUTHOR]

The COVID-19 pandemic resulted in profound disruption to geography higher education. A pivot to online teaching required staff to rapidly adapt their practices to novel digital spaces. Whilst many studies have reported the different pedagogic approaches adopted, fewer have evaluated the resultant student learning experience. In this study, we aimed to create an evidence base regarding the benefits and challenges of online learning during the pandemic, mapped against the teaching, cognitive and social presences of the Community of Inquiry framework. We adopted a mixed-methods approach of online surveys (105 students) and focus groups (14 students), undertaken across two undergraduate geography programmes in a British university, exploring the benefits of asynchronous and synchronous online learning, and assessment and feedback strategies. We discovered flexibility in student work patterns and use of technology to facilitate engagement in learning. We also identified key challenges for students such as time management, maintaining motivation, engaging in online classes, and feeling part of an online learning community. We identify best practice in collaborative-constructivist online learning, so that in the event of any future remote pivot, or with sustained adoption of blended modes of delivery, we can achieve a high-quality student learning experience. [ABSTRACT FROM AUTHOR]

The t‐distributed stochastic neighbour embedding algorithm or t‐SNE is a non‐linear dimension reduction method used to visualise multivariate data. It enables a high‐dimensional dataset, such as a set of infrared spectra, to be represented on a single, typically two‐dimensional graph, revealing its global and local structure. t‐SNE is very popular in the machine learning community and has been applied in many fields, generally with the aim of visualising large datasets. In vibrational spectroscopy, t‐SNE is gaining notoriety but principal component analysis (PCA) remains by far the reference method for exploratory analysis and dimension reduction. However, t‐SNE may represent a real aid in the analysis of vibrational spectroscopic datasets. It provides an at‐a‐glance global view of the dataset allowing to distinguish the main factors influencing the spectral signal and the hierarchy between these factors, and gives an indication on the possibility of performing predictive modelling. It can also provide great support in the choice of the pre‐processing, by comparing rapidly different general pre‐processing approaches according to their effect on the variable of interest. Here we propose to illustrate these advantages using different datasets. We also propose an approach based on a synergy between the t‐SNE and PCA methods, allowing respective advantages of each to be exploited. The t‐distributed stochastic neighbour embedding algorithm, or t‐SNE, is a nonlinear method enabling the visualisation of multivariate datasets in lower dimensions. In vibrational spectroscopy, t‐SNE can provide a rapid global overview of a dataset and its influencing factors, assess the potential for predictive modelling and aid with the selection of the pre‐processing workflow. These advantages are illustrated with two real vibrational spectroscopic datasets. A data exploration approach based on the synergy between t‐SNE and PCA is also proposed. [ABSTRACT FROM AUTHOR]

This paper reports on a study that explored the user-driven inter-organisational professional learning that teachers from two education systems collectively engaged in via social media. A dedicated Facebook group was established to enable Australian early childhood teachers (sector one) and primary school teachers (sector two) to engage in collaborations that would support children's transition to school. Using an Activity Theory framework, findings showed that the site was mainly used by early childhood teachers to seek peer support in meeting reporting requirements, and that school teachers rarely posted. As such, the capacity of the Facebook group to support inter-organisational cross-sector collaboration and learning was challenged as determined by the most active participants within the site, while reinforcing the group's capacity to support 'just in time' intra-organisational professional learning. This highlights a need for further research to explore the ways and means by which social media may best facilitate cross-sector collaboration between education systems, such as a more focussed and integrated use of social media during face-to-face cross-sector professional learning. [ABSTRACT FROM AUTHOR]

Recruiting the public to participate in the scientific process can be invaluable in furthering our understanding of global environmental change. Several long‐term citizen science projects have been active for over a decade, with most involving the public in the data collection phase of the scientific process. Our team has recently inherited a long‐term citizen science project called the Community Lake Ice Collaboration (CLIC). We have benefited from engaging with a community of experienced participants who have collected reliable data for hundreds of lakes across the USA for over 30 yr. Collecting data at this scale would have been logistically and financially challenging without participants volunteering their time and effort. Here, we aim to provide an overview of the lessons we have learned from CLIC and other citizen science projects and develop nine guidelines for establishing and sustaining a long‐term citizen science project. [ABSTRACT FROM AUTHOR]

Background: Transfer students are a growing population of undergraduate psychology majors who are academically at-risk but often underserved by 4-year institutions. Objective: We describe a transfer learning community (TLC) for entering psychology transfer students and report its impact on participants social and academic integration and success. Method: A mixed-method design was used to compare academic success metrics (archival measures of first-semester grade point average and probation rates) of participants with entering transfer majors not participating in the learning community. A participant survey was also created and used to measure student perceptions of the TLC, as well as perceived challenges of transferring during the Covid-19 pandemic and how the TLC helped to mitigate them. Results: During its first year, the TLC served 72 entering transfer majors across two cohorts. Participants significantly out-performed non-participant psychology transfer majors on first-semester grade point average and were significantly less likely to be placed on academic probation after their first semester than non-participant transfer majors. Conclusion: The widely-reported benefits of learning communities for first-year psychology students are adaptable to the more at-risk population of transfer students. Teaching Implications: Technology-enabled teaching tools can be used to build virtual community in hybrid and synchronous online learning environments. [ABSTRACT FROM AUTHOR]

Bayesian neural networks (BNNs) have recently regained a significant amount of attention in the deep learning community due to the development of scalable approximate Bayesian inference techniques. There are several advantages of using a Bayesian approach: parameter and prediction uncertainties become easily available, facilitating more rigorous statistical analysis. Furthermore, prior knowledge can be incorporated. However, the construction of scalable techniques that combine both structural and parameter uncertainty remains a challenge. In this paper, we apply the concept of model uncertainty as a framework for structural learning in BNNs and, hence, make inferences in the joint space of structures/models and parameters. Moreover, we suggest an adaptation of a scalable variational inference approach with reparametrization of marginal inclusion probabilities to incorporate the model space constraints. Experimental results on a range of benchmark datasets show that we obtain comparable accuracy results with the competing models, but based on methods that are much more sparse than ordinary BNNs. [ABSTRACT FROM AUTHOR]

This article, titled "Lessons Learned from a Community Water Garden," discusses the creation of a water garden in Carroll Creek, Frederick, MD, USA, in response to frequent algal blooms in the creek. The project was undertaken by a water gardening community with limited funds and knowledge of aquaculture. The water garden aimed to mitigate the algae problem by providing shade and covering 60% of the water surface with lotus, water lilies, and bog plants. The project successfully attracted volunteers and funding from local civic organizations and has become a popular attraction, contributing to the economic vitality of the area. The article emphasizes the importance of community engagement and the power of vision in achieving successful projects. [Extracted from the article]

This article explores the evolutionary nature of higher education and how it continues to stress the need for the professoriate to be active within scholarly teaching. This qualitative study examines how each author navigates SoTL research and analyzes the authors' reflective journaling for themes and sub‐themes for alignment with Scholarship Reconsidered, Scholarship Assessed and Community of Scholars models. Study findings support the existing literature and showcases the importance of establishing a community of scholars that offers its members a place of psychological safety and support. [ABSTRACT FROM AUTHOR]

Efficient management of renewable energy resources is imperative for promoting environmental sustainability and optimizing the utilization of clean energy sources. This paper presents a pioneering European-scale study on energy management within renewable energy communities (RECs). With a primary focus on enhancing the social welfare of the community, we introduce a reinforcement learning (RL) controller designed to strategically manage Battery Energy Storage Systems (BESSs) and orchestrate energy flows. This research transcends geographical boundaries by conducting an extended analysis of various energy communities and diverse energy markets across Europe, encompassing different regions of Italy. Our methodology involves the implementation of an RL controller, leveraging optimal control theory for training and utilizing only real-time data available at the current time step during the test phase. Through simulations conducted in diverse contexts, we demonstrate the superior performance of our RL agent compared to a state-of-the-art rule-based controller. The agent exhibits remarkable adaptability to various scenarios, consistently surpassing existing rule-based controllers. Notably, we illustrate that our approach aligns with the intricate patterns observed in both Italian and European energy markets, achieving performance levels comparable to an optimal controller assuming perfect theoretical knowledge of future data. [ABSTRACT FROM AUTHOR]

Only a few states have adopted a paid family leave (PFL) policy in the United States of America. Local media described the 2019 Oregon PFL legislation as "the most progressive" policy in the country, with coalitions as crucial policy advocates. This case study applies the Advocacy Coalition Framework (ACF) to examine policy learning and negotiated agreements as causal mechanisms to explain the adoption of the PFL. We identified three modes of policy learning: previous policy cycles, learning from other coalitions, and learning from community organizations. ACF explains the evolution of negotiated agreements based on the stability of coalition belief systems, including consensus on leave time, inclusivity, and cost-sharing contributions. ACF helps describe how coalitions adopted progressive ideas such as equity. However, ACF's elements that allow a deeper exploration of narratives were missing. Future studies should include interviews with coalition members and compare state policies to assess strategies. Future policy initiatives could integrate feedback from community organizations into policy strategies. [ABSTRACT FROM AUTHOR]

Handling large datasets and calculating complex statistics on huge datasets require important computing resources. Using subsampling methods to calculate statistics of interest on small samples is often used in practice to reduce computational complexity, for instance using the divide and conquer strategy. In this article, we recall some results on subsampling distributions and derive a precise rate of convergence for these quantities and the corresponding quantiles. We also develop some standardisation techniques based on subsampling unstandardised statistics in the framework of large datasets. It is argued that using several subsampling distributions with different subsampling sizes brings a lot of information on the behaviour of statistical learning procedures: subsampling allows to estimate the rate of convergence of different algorithms, to estimate the variability of complex statistics, to estimate confidence intervals for out-of-sample errors and interpolate their values at larger scales. These results are illustrated on simulations, but also on two important datasets, frequently analysed in the statistical learning community, EMNIST (recognition of digits) and VeReMi (analysis of Network Vehicular Reference Misbehavior). [ABSTRACT FROM AUTHOR]

Minimax problems have gained tremendous attentions across the optimization and machine learning community recently. In this paper, we introduce a new quasi-Newton method for the minimax problems, which we call J-symmetric quasi-Newton method. The method is obtained by exploiting the J-symmetric structure of the second-order derivative of the objective function in minimax problem. We show that the Hessian estimation (as well as its inverse) can be updated by a rank-2 operation, and it turns out that the update rule is a natural generalization of the classic Powell symmetric Broyden method from minimization problems to minimax problems. In theory, we show that our proposed quasi-Newton algorithm enjoys local Q-superlinear convergence to a desirable solution under standard regularity conditions. Furthermore, we introduce a trust-region variant of the algorithm that enjoys global R-superlinear convergence. Finally, we present numerical experiments that verify our theory and show the effectiveness of our proposed algorithms compared to Broyden's method and the extragradient method on three classes of minimax problems. [ABSTRACT FROM AUTHOR]

In addition to learning effectively and effectively in online learning communities (OLCs), students must be in a secure environment and privacy must be respected. The study aimed to identify privacy violations that university students encounter in OLCs, and identify recommendations and some strategies to implement to ensure privacy at the OLC. A qualitative research method was applied in the study. First, focus group interviews were conducted with a total of 108 students who participated actively in the OLC. After analysing the interview data, a privacy violations report (RPV‐OLC) was prepared at the OLC. According to the RPV‐OLC, a problem description for the Delphi technique was designed and 17 experts were identified for the interview. Delphi panels with experts were completed in 5 months. The content analysis process was used for the data obtained. Students said they encountered information sharing and access, access to information, system/course settings, resistance to negotiations, communication problems and lack of knowledge. On the other hand, experts suggest that different suggestions and strategies could be implemented to ensure privacy in the OLC, such as partial privacy, mutual trust, contextual privacy agreements, role‐based or anonymous identity use, and privacy education. The study focused on defining the concept of privacy and student privacy violations in online learning environments and providing solutions for educators, students and system administrators. Practitioner notesWhat is already known about this topicDigital learning environments pose a threat to the privacy of students.The fact that privacy violations in the OLC can trigger concerns among students has a negative impact on learning processes.There is high interest at the moment in how to plan the most effective teaching in online learning environments—but privacy is being neglected.What this paper addsThis article explains the privacy violations university students encounter at the OLC.This document contains guidelines on how to mitigate and prevent privacy violations in OLC.The results of the study also provide suggestions for privacy protection at the OLC.Implications for practice and/or policyIt is important that the instructors who participate in the OLC are informed about the necessary measures, rules, principles, activities and follow‐up measures to protect their personal privacy.If we can overcome students' concerns about privacy violations at the OLC, we can create more comfortable learning spaces.The aim of learning should not only be to peak performance; the most important responsibility of educators or system administrators is to preserve each student's private space, while at the same time achieving the best performance. [ABSTRACT FROM AUTHOR]

Social and economic participation is increasingly dependent on the proficient use of digital technologies in everyday life. Digital skills, alongside literacy and numeracy, are now foundational skills that all Australians should have. Yet, how policy makers, scholars, and adult educators can effectively identify people's current and desired digital skills is not well established. Existing frameworks tend to be instrumentalist, listing generic abilities that all citizens need for success. Digital abilities, however, are context-specific and attempts to generalise need to be responsive to local circumstances. We present a theory-driven Digital Ability Self-Assessment Tool, developed for a government agency and piloted in a community organisation. The Tool takes a distinctive socio-material approach to overcome shortcomings of other digital ability assessment tools, which often focus on normative measures of skills attainment. Overall, we argue that authentic digital skills policy and programme development must account for the contexts in which digital technologies are used. [ABSTRACT FROM AUTHOR]

University and college leaders are tasked with enhancing student outcomes with fewer resources. Student retention is one such key outcome of interest for many policy makers as well as for university administrators. Over the years, administrators have turned to High Impact Practices (HIPs) such as Learning Communities (LCs) to aid in retention. This quantitative study explores the impact LCs have on student retention at a large R1 university in the Midwest. Additionally, the financial return on investment in LCs at this institution is measured via tuition dollars generated from students who are retained as a result of their participation in a LC at the institution. Two key findings of this study are that LCs are positively associated with increased odds of student retention, and that investing in LCs makes good financial sense. Our research contributes to the scholarship on retention attributable to LCs and provides researchers and practitioners with a "template" to evaluate the efficacy of specific retention initiatives in relation to their financial return on investments. [ABSTRACT FROM AUTHOR]

This article considers the possibilities and limits of reimagining international development policy by taking the values, practices, and worldviews of Indigenous communities as its starting point. Drawing on ethnographic research in Guatemala, we contrast the development industry's overwhelming focus on economic growth as the gold standard of well-being with the perspective of Maya-Achí groups, who insist that growth and modernization must not come at the expense of the ecology, food sovereignty, or Indigenous ways of life connected to the land. We argue that the Maya-Achi organizations with whom we collaborate offer a philosophy and practice better attuned to the urgency of the climate crisis than that of the dominant model of development. To bring the international development agenda in line with local climate action, we propose reconceiving Development as Buen Vivir—an Indigenous philosophy of good living. To do so, we propose three lines of action: (1) Increasing Funding for Indigenous-led climate action; (2) Re-conceptualizing development practices to align with Buen Vivir, and (3) Transforming social and economic policies. [ABSTRACT FROM AUTHOR]

Introduction: In the post-epidemic era, blended learning has become a social trend for the future of higher education, and scholars have endeavored to understand the factors that influence student learning in these blended communities. Communities of Inquiry is a conceptual framework that describes the components of blended learning environments, indicating teaching presence, social presence, and cognitive presence. However, the framework fails to adequately explore how individual learning motivational factors influence student learning. Therefore, this study extends the Community of Inquiry framework by drawing on a positive psychological construct-academic buoyancy to reveal the relationship between academic buoyancy and the three presences through empirical research. Methods: The theoreticalmodel was validated by SPSS 26.0 and smartPLS4.0. To evaluate the measurement and structural models, structural equation modeling (SEM) was carried out using the partial least squares (PLS) method. Findings: (a) Teaching presence positively predicts academic buoyancy, and academic buoyancy positively predicts social presence and cognitive presence; (b) academic buoyancy mediates teaching presence and social presence, as well as teaching presence and cognitive presence; and (c) academic buoyancy acts as a chainmediator between teaching presence and cognitive presence through social presence. Discussion: The results of this study fill a gap in the multiple roles of individual positive psychological construct-academic buoyancy in blended learning communities, extend the Community of Inquiry theoretical framework, and provide empirical evidence for blended learning quality and practical improvement strategies. [ABSTRACT FROM AUTHOR]

Although e-Portfolio is acknowledged as one of the powerful pedagogical practices that enhance learning in higher education (HE), not much is known about the types of social media (SM) utilized as e-Portfolios and the benefits for students. This literature analysis, using directed content analysis, aims to explore the above vacuum. The research questions in this study are: (1) In what ways do the SM as e-Portfolios benefit students in HE? (2) To what extent are the benefits of SM as e-Portfolios comparable to those of conventional e-Portfolios? and (3) What are the drawbacks that practitioners and researchers need to be concerned with? Findings indicate that blogs are the most popular SM used as e-Portfolios to support learning, followed by social networking sites and collaborative projects. The study yields 13 advantages and 12 drawbacks when SM is manipulated as e-Portfolios. These findings conclude that the use of SM as e-Portfolios has a great potential in supporting students' learning and development by providing an environment for them to learn meaningfully from their experiences and engage in critical reflections and dialogues that allow them to gain new knowledge and valuable insights and thus, improve their skills. A pedagogical framework for the planning and implementation of SM as e-Portfolios is suggested based on the findings and aims of the papers that were reviewed. [ABSTRACT FROM AUTHOR]

The aim is to investigate the effects of delivering group-based personalized teaching via an electronic performance support system (EPSS) in an online medical informatics course on medical students' academic achievement and community of inquiry levels. The basic working principle of EPSS is to provide the most appropriate teaching methods to the educator, according to the students' variables and learning outcomes. The research design of this study involved a quasi-experimental pretest-posttest design with a control group and the qualitative research method of answers to follow-up questions. While the students in the experimental group were taught various teaching approaches (flipped classroom, brain storming, case-based learning, debate and think-pair-share), the students in the control group were taught directly by online lessons according to direct instruction method. It was seen that the experimental procedure had a positive effect on both the academic achievement and community of inquiry levels of the experimental group compared to the control group. Besides, students in the experimental group reported that the lessons were more effective when they were enjoyable, motivating, focused on discussion and research, and were taught through cases. The group personalized teaching method used in this study could be a guide for future studies. [ABSTRACT FROM AUTHOR]

In this study, we collaborated with the National Library of Korea's Library Education Institute to pilot the operation of a librarian learning community and aimed to measure its effectiveness based on the improvement of librarian competencies. A total of 28 participants attended four programs, and the research results are as follows: First, when analyzing competencies in terms of community capacity, attitude competency was the highest, followed by knowledge and skills competencies in that order. The correlation analysis among competencies showed that there were mutual correlations among the three competencies. Second, in the pre-post analysis of the three competencies and their sub-competencies, it was found that all competencies, including sub-competencies, improved. Based on the research results, it was confirmed that through the operation of the librarian learning community, the average competencies of librarians can be improved, which can directly impact the enhancement of the library's service capabilities. Therefore, it is believed that there is a need to activate the operation of librarian learning communities and explore nationwide dissemination strategies for the operating model. [ABSTRACT FROM AUTHOR]

Retrieval practice is an evidence‐based approach to teaching; here, we evaluate the use of PeerWise for embedding retrieval practice into summative assessment. PeerWise allows anonymous authoring, sharing, answering, rating, and feedback on peer‐authored multiple choice questions. PeerWise was embedded as a summative assessment in a large first‐year introductory biochemistry module. Engagement with five aspects of the tool was evaluated against student performance in coursework, exam, and overall module outcome. Results indicated a weak‐to‐moderate positive but significant correlation between engagement with PeerWise and assessment performance. Student feedback showed PeerWise had a polarizing effect; the majority recognized the benefits as a learning and revision tool, but a minority strongly disliked it, complaining of a lack of academic moderation and irrelevant questions unrelated to the module. PeerWise can be considered a helpful learning tool for some students and a means of embedding retrieval practice into summative assessment. [ABSTRACT FROM AUTHOR]

Responding to contemporary global challenges will require quality education, which presupposes changes in schools' organizational structures, new roles and mentalities for the players involved and a culture of learning at the different levels of organizations. In this regard, we strive to understand how changing certain organizational structures, more specifically the organization of teaching by educational teams operating as professional learning communities, can have an impact on individual, collective and organizational learning at schools. To this end, we have adopted a qualitative research paradigm, put into operation through a case study, based on a qualitative-quantitative approach. We combined a descriptive statistical approach consisting of two questionnaires, which have been decoded and interpreted both structurally and semantically, with a content analysis of interviews, focused discussion groups and field diary notes, to examine a number of unique contexts and the perspectives of individual players. We concluded that a change in organizational structures is an essential but insufficient condition. What is needed is a change in the players' beliefs and collaboration that provides deep learning. As such, a substantial change in education at the level of educational organization requires joint action at the levels of structure, middle leadership, beliefs and professional cultures. [ABSTRACT FROM AUTHOR]

Studying determinant factors for effective learning communities has become a priority for sustainable education in Yemen. Yet, comprehensive knowledge of academic self-efficacy (ASE), academic commitment (AC), and learning outcomes (LOs) variables is limited in higher education, particularly. This research aimed to examine the role of these variables in enhancing the learning attainments of Yemeni university students by studying the direct impacts of ASE and AC on LOs, the influence of ASE on AC, and, lastly, how AC mediates the relationship between ASE and LOs. The research employed a meticulously validated measurement model encompassing 12 ASE items, 10 LOs items, and 27 AC items, using robust statistical methods such as EFA, CFA, and SEM. The study engaged 611 students aged 19–27. Empirical findings confirmed the significant impact of ASE on AC and LOs, as well as the impact of AC on LOs, both of which serve as vital factors in the academic context. Moreover, the research identified an indirect influence of ASE on LOs mediated through AC. On the other hand, the results concerning the demographic variables indicated significant differences between gender and age groups, as well as between age and educational level groups. However, no differences were found when comparing gender and level groups or when considering gender, age, and level together for ASE and LOs. In contrast, for students' AC, all comparisons were significant except for the gender and age comparison. The study offers valuable insights into educational practices and policies and concludes with practical recommendations to enhance the academic environment, informed by these study outcomes. [ABSTRACT FROM AUTHOR]

This paper explores the formation of a learning community facilitated by custom collaborative learning software. Drawing on research in group cognition, knowledge building discourse, and learning analytics, we conducted a mixed-methods field study involving an asynchronous online discussion consisting of 259 messages posted by 50 participants. The cluster analysis results provide evidence that the recommender system within the software can support the formation of a learning community with a small peripheral cluster. Regarding knowledge building discourse, we identified the distinct roles of central, intermediate (i.e., middle of three clusters), and peripheral clusters within a learning community. Furthermore, we found that message lexical complexity does not correlate to the stages of knowledge building. Overall, this study contributes to the group cognition theory to deepen our understanding about collaboration to construct new knowledge in online discussions. Moreover, we add a much-needed text mining perspective to the qualitative interaction analysis model. [ABSTRACT FROM AUTHOR]

Background: Research clerkships are usually designed as individual learning projects focusing on research skills training, such as research design, data analysis and reporting. When the COVID‐19 pandemic triggered an urgent need for digital education, we redesigned a research clerkship with the challenging aim to maintain original quality for more students than usual with limited teaching staff. Approach: We introduced the concept of a research learning community (RLC) with co‐teaching and co‐learning to a group of 14 students and seven teaching faculty using digital platforms. Small groups of students participated in the RLC, which was supervised weekly by the teachers. Research experts were continuously involved and led workshops. Evaluation: Using a qualitative design, we analysed experiences from the perspectives of students and faculty. We performed an inductive thematic content analysis of three focus group interviews and used 14 student reports for triangulation. The results indicate that apart from developing research skills, students valued peer assistance, attention to uncertainty and learning beyond individual research projects. The teachers/research experts reported that co‐teaching and co‐learning had contributed to their professional development. In terms of organisation, students and faculty recognised that the RLC model allowed for interdisciplinary learning, facilitated by a digital platform. Implications: To develop students' research skills, embedding a clerkship in an RLC is an attractive alternative to individual research clerkships. The obligatory learning goals are met. Co‐learning and co‐teaching foster faculty's and students' professional development. When translating to other curricula, we recommend stating common goals in addition to individual objectives. [ABSTRACT FROM AUTHOR]

Environments or learning spaces in school classrooms contribute to the improvement of educational processes. Specifically, in the Early Childhood Education, teachers can design learning spaces, which fosters an environment of safety and autonomy for the students of the Early Childhood Education. This study explores the teachers' narratives concerning the meanings they assign to the school space, the criteria by which they design and transform it, and its connection to model of educational. We have collected 34 stories of pre-primary education teachers through a semi-structured interview in which they had to reflect on this pedagogical design. As most relevant result we identify a model sustained under constructivist or cognitive criteria, rather than a dialogical one. We conclude that this epistemological perspective of the construction of the school environment should be oriented towards a more participatory and co-constructed model of the educational space.

With the rapid development of international communication, the number of English courses has shown an explosive growth trend, which has caused a serious problem of information overload, resulting in poor teaching performance of recommended English courses. To solve this problem, this paper proposes a graph convolutional neural network model based on College English course texts, students’ major, English foundation and network structure characteristics. First, by analyzing the relevant data of College English courses and combining with graph neural network, an English course recommendation algorithm model based on the College English learning strategy of proximity comparison is proposed. Then, the College English texts are taken as feature input, and multi-layer graph convolutional neural network is used to process the above graph neural network structure. Attention mechanism is introduced to enhance the representation of graph features in College English skills. Finally, multi-layer attention model is used to process the courses that users have learned, and intelligent course recommendation is made by combining the multi-layer attention modeling of College English skills. The experimental data show that the proposed method achieves the best performance compared with the commonly used College English course recommendation method.

There is a gap between current methods that explain deep learning models that work on imbalanced image data and the needs of the imbalanced learning community. Existing methods that explain imbalanced data are geared toward binary classification, single layer machine learning models and low dimensional data. Current eXplainable Artificial Intelligence (XAI) techniques for vision data mainly focus on mapping predictions of specific instances to inputs, instead of examining global data properties and complexities of entire classes. Therefore, there is a need for a framework that is tailored to modern deep networks, that incorporates large, high dimensional, multi-class datasets, and uncovers data complexities commonly found in imbalanced data. We propose a set of techniques that can be used by both deep learning model users to identify, visualize and understand class prototypes, sub-concepts and outlier instances; and by imbalanced learning algorithm developers to detect features and class exemplars that are key to model performance. The components of our framework can be applied sequentially in their entirety or individually, making it fully flexible to the user's specific needs (https://github.com/dd1github/XAI_for_Imbalanced_Learning). [ABSTRACT FROM AUTHOR]

Recently, transfer learning has gained popularity in the machine learning community. Transfer Learning (TL) has emerged as a promising paradigm that leverages knowledge learned from one or more related domains to improve prediction accuracy in a target domain with limited data. However, for time series forecasting (TSF) applications, transfer learning is relatively new. This paper addresses the need for empirical studies as identified in recent reviews advocating the need for practical guidelines for Transfer Learning approaches and method designs for time series forecasting. The main contribution of this paper is the suggestion of a comprehensive framework for Transfer Learning Sensitivity Analysis (SA) for time series forecasting. We achieve this by identifying various parameters seen from various angles of transfer learning applied to time series, aiming to uncover factors and insights that influence the performance of transfer learning in time series forecasting. Undoubtedly, symmetry appears to be a core aspect in the consideration of these factors and insights. A further contribution is the introduction of four TL performance metrics encompassed in our framework. These TL performance metrics provide insight into the extent of the transferability between the source and the target domains. Analyzing whether the benefits of transferred knowledge are equally or unequally accessible and applicable across different domains or tasks speaks to the requirement of symmetry or asymmetry in transfer learning. Moreover, these TL performance metrics inform on the possibility of the occurrence of negative transfers and also provide insight into the possible vulnerability of the network to catastrophic forgetting. Finally, we discuss a sensitivity analysis of an Ensemble TL technique use case (with Multilayer Perceptron models) as a proof of concept to validate the suggested framework. While the results from the experiments offer empirical insights into various parameters that impact the transfer learning gain, they also raise the question of network dimensioning requirements when designing, specifically, a neural network for transfer learning. [ABSTRACT FROM AUTHOR]

Deep neural network ensembles combine the wisdom of multiple deep neural networks to improve the generalizability and robustness over individual networks. It has gained increasing popularity to study and apply deep ensemble techniques in the deep learning community. Some mission-critical applications utilize a large number of deep neural networks to form deep ensembles to achieve desired accuracy and resilience, which introduces high time and space costs for ensemble execution. However, it still remains a critical challenge whether a small subset of the entire deep ensemble can achieve the same or better generalizability and how to effectively identify these small deep ensembles for improving the space and time efficiency of ensemble execution. This article presents a novel deep ensemble pruning approach, which can efficiently identify smaller deep ensembles and provide higher ensemble accuracy than the entire deep ensemble of a large number of member networks. Our hierarchical ensemble pruning approach (HQ) leverages three novel ensemble pruning techniques. First, we show that the focal ensemble diversity metrics can accurately capture the complementary capacity of the member networks of an ensemble team, which can guide ensemble pruning. Second, we design a focal ensemble diversity based hierarchical pruning approach, which will iteratively find high quality deep ensembles with low cost and high accuracy. Third, we develop a focal diversity consensus method to integrate multiple focal diversity metrics to refine ensemble pruning results, where smaller deep ensembles can be effectively identified to offer high accuracy, high robustness and high ensemble execution efficiency. Evaluated using popular benchmark datasets, we demonstrate that the proposed hierarchical ensemble pruning approach can effectively identify high quality deep ensembles with better classification generalizability while being more time and space efficient in ensemble decision making. We have released the source codes on GitHub at https://github.com/git-disl/HQ-Ensemble. [ABSTRACT FROM AUTHOR]

Multi-label classification problem, where each instance can be associated with multiple labels, has received considerable attention from machine learning community. To address the inherent challenges of multi-label classification including data imbalance, label dependence, and high dimensionality, ensemble approaches have been developed, gaining popularity across various real-world applications. This paper proposes a novel ensemble method called ConfBoost that addresses these challenges and enhances the generalization ability of learning systems. ConfBoost which is a meta-model based on a weighted stacking paradigm using local confidence, combines heterogeneous and complementary ensembles of multi-label classifiers. The proposed approach achieves two main objectives: Firstly, by focusing on label weights based on their confidence scores, the model can generate more relevant predictions and enhance the accuracy at the base-level by mitigating the impact of irrelevant labels during the stacking process. Moreover, assigning higher weights to certain labels exhibits better discrimination and adaptability to capture complex label relationships. Second, applying adjusted thresholds enables the model to generate predictions adapted to the specific characteristics of each label, effectively addressing imbalanced label distributions. Extensive experiments on publicly available datasets demonstrate that ConfBoost outperforms conventional combination methods and consistently surpasses related state-of-the-art methods. These findings highlight the effectiveness and potential of ConfBoost as an advanced ensemble method for multi-label classification tasks. [ABSTRACT FROM AUTHOR]

A fundamental problem in functional data analysis is to classify a functional observation based on training data. The application of functional data classification has gained immense popularity and utility across a wide array of disciplines, encompassing biology, engineering, environmental science, medical science, neurology, social science, and beyond. The phenomenal growth of the application of functional data classification indicates the urgent need for a systematic approach to develop efficient classification methods and scalable algorithmic implementations. Therefore, we here conduct a comprehensive review of classification methods for functional data. The review aims to bridge the gap between the functional data analysis community and the machine learning community, and to intrigue new principles for functional data classification. This article is categorized under:Statistical Learning and Exploratory Methods of the Data Sciences > Clustering and ClassificationStatistical Models > Classification ModelsData: Types and Structure > Time Series, Stochastic Processes, and Functional Data [ABSTRACT FROM AUTHOR]

Representing mixing state of black carbon (BC) is challenging for global climate models (GCMs). The Community Atmosphere Model version 6 (CAM6) with the four‐mode version of the Modal Aerosol Module (MAM4) represents aerosols as fully internal mixtures with uniform composition within each aerosol mode, resulting in high degree of internal mixing of BC with non‐BC species and large mass ratio of coating to BC (RBC, the mass ratio of non‐BC species to BC in BC‐containing particles). To improve BC mixing state representation, we coupled a machine learning (ML) model of BC mixing state index trained on particle‐resolved simulations to the CAM6 with MAM4 (MAM4‐ML). In MAM4‐ML, we use RBC to partition accumulation mode particles into two new modes, BC‐free particles and BC‐containing particles. We adjust RBC to make the modeled BC mixing state index (χmode) match the one predicted by the ML model (χML). On a global average, the mass fraction of BC‐containing particles in accumulation mode decreases from 100% (MAM4‐default) to 48% (MAM4‐ML). The globally averaged χmode decreases from 78% (MAM4‐default) to 63% (MAM4‐ML, 19% reduction) and agrees well with χML (66%). The RBC decreases by 52% for accumulation mode and better agrees with observations. The hygroscopicity drops by 9% for BC‐containing particles in accumulation mode, leading to a 20% reduction in the BC activation fraction. The surface BC concentration increases most (6.9%) in the Arctic, and the BC burden increases by 4%, globally. Our study highlights the application of the ML model for improving key aerosol processes in GCMs. Plain Language Summary: Black carbon (BC) mixing state is an important property for assessing impacts of BC on human health and climate. It describes how BC is distributed among the particle population and within individual particles. Due to computational costs, global climate models (GCMs) consider aerosols as either external mixtures (e.g., bulk models) or fully internal mixtures with uniform composition within a certain size range (e.g., an aerosol bin in sectional models or an aerosol mode in modal models). To improve BC mixing state representation, we coupled a machine learning (ML) model learned from particle‐resolved simulations to the online simulations of the Community Atmosphere Model version 6 (CAM6) with the four‐mode version of the Modal Aerosol Module (MAM4‐ML). We used the mass ratio of coating to BC (RBC) to determine the mass fraction of BC‐free particles and BC‐containing particles, and we adjusted RBC to make the modeled BC mixing state match ML model. The MAM4‐ML model reduces the degree of internal mixture of BC substantially. In addition, MAM4‐ML RBC is reduced by 52%, leading to a 20% reduction in BC activation fraction and a 4% increase in BC burden. This study proposes a method for the application of the ML to improve GCMs. Key Points: Machine learning model trained on particle‐resolved simulations has been coupled to CAM6 MAM4 to improve BC mixing state representationThe BC mixing state index is reduced by 19%, and the mass ratio of coating to BC is reduced by 52%, in better agreement with observationsDue to the drop in hygroscopicity of BC‐containing aerosol (9%), BC activation fraction decreased by 20%, and BC burden increased by 4% [ABSTRACT FROM AUTHOR]

The most effective and beneficial learning communities are ones that focus on children's wellness as well as their academics. [ABSTRACT FROM AUTHOR]

This article illustrates how two teachers, a general educator and a special educator, used an interactive process in planning an interdisciplinary science unit for an inclusive Kindergarten classroom and how they addressed Universal Design for Learning (UDL) components. This planning process included capitalizing on each teacher's knowledge and expertise, layering resources from within their learning community iteratively over time (not a one‐meeting‐and‐done effort), and harnessing the power of multimodal, interdisciplinary learning. This iterative process was key in allowing teachers to evaluate student engagement and layer in UDL components to address evolving students' needs. [ABSTRACT FROM AUTHOR]

We used a phenomenological research design to explore the impact of a Faculty Learning Community (FLC) to support STEM faculty members' development of cultural wealth principles and teaching practices. In this study, we analyzed teaching philosophy statements to learn how the FLC shaped STEM faculty participants' expression of teaching beliefs and values. Because effective teaching philosophy statements are linked to good teaching and transparently designed assignments, the language used to articulate teaching values can inform how we build equity-minded and culturally relevant educational spaces. Findings demonstrated that some faculty members developed and articulated teaching beliefs and intended practices that were consistent with culturally relevant teaching and cultural wealth. However, the study did not discover evidence of practice intentions that reflected these culturally relevant pedagogical principles related to the specific assessment of student learning. We discuss implications for faculty developers at HSIs to support STEM faculty members' use of culturally relevant and community engaged concepts and teaching approaches. [ABSTRACT FROM AUTHOR]

Partially linear additive models (PLAMs) have attracted much attention in the statistical machine learning community due to their interpretability and flexibility in data-driven prediction and inference. Since the performance of PLAMs is closely related to the structure information of linear and nonlinear components, several approaches have been proposed for regression estimation and data-driven structure discovery. However, the existing automatic discovery strategy is limited to the mean regression framework and is usually sensitive to non-Gaussian noises, e.g. skewed noise and heavy-tailed noise. To further improve the robustness of PLAMs, this paper proposes a Robust Partially Linear Trend Filtering (RPLTF) for regression estimation and structure discovery by integrating the mode-induced error metric and the trend filtering-based nonlinear approximation into regularized PLAMs. Besides the computing algorithm of RPLTF, we establish its upper bound on generalization error in theory. Empirical examples are provided to validate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Statistical prediction models have gained popularity in applied research. One challenge is the transfer of the prediction model to a different population which may be structurally different from the model for which it has been developed. An adaptation to the new population can be achieved by calibrating the model to the characteristics of the target population, for which numerous calibration techniques exist. In view of this diversity, we performed a systematic evaluation of various popular calibration approaches used by the statistical and the machine learning communities for estimating two‐class probabilities. In this work, we first provide a review of the literature and, second, present the results of a comprehensive simulation study. The calibration approaches are compared with respect to their empirical properties and relationships, their ability to generalize precise probability estimates to external populations and their availability in terms of easy‐to‐use software implementations. Third, we provide code from real data analysis allowing its application by researchers. Logistic calibration and beta calibration, which estimate an intercept plus one and two slope parameters, respectively, consistently showed the best results in the simulation studies. Calibration on logit transformed probability estimates generally outperformed calibration methods on nontransformed estimates. In case of structural differences between training and validation data, re‐estimation of the entire prediction model should be outweighted against sample size of the validation data. We recommend regression‐based calibration approaches using transformed probability estimates, where at least one slope is estimated in addition to an intercept for updating probability estimates in validation studies. [ABSTRACT FROM AUTHOR]

Recurrent neural networks are effective models to process sequences. However, they are unable to learn longterm dependencies because of their inherent sequential nature. As a solution, Vaswani et al. introduced the Transformer, a model solely based on the attention mechanism that is able to relate any two positions of the input sequence, hence modelling arbitrary long dependencies. The Transformer has improved the state-of-the-art across numerous sequence modelling tasks. However, its effectiveness comes at the expense of a quadratic computational and memory complexity with respect to the sequence length, hindering its adoption. Fortunately, the deep learning community has always been interested in improving the models' efficiency, leading to a plethora of solutions such as parameter sharing, pruning, mixed-precision, and knowledge distillation. Recently, researchers have directly addressed the Transformer's limitation by designing lower-complexity alternatives such as the Longformer, Reformer, Linformer, and Performer. However, due to the wide range of solutions, it has become challenging for researchers and practitioners to determine which methods to apply in practice to meet the desired tradeoff between capacity, computation, and memory. This survey addresses this issue by investigating popular approaches to make Transformers faster and lighter and by providing a comprehensive explanation of the methods' strengths, limitations, and underlying assumptions. [ABSTRACT FROM AUTHOR]

The book "Inside Jewish Day Schools: Leadership, Learning and Community" by Alex Pomson and Jack Wertheimer provides a comprehensive examination of Jewish day schools in North America. The authors aim to capture the experiences and challenges faced by these schools and their leaders, offering valuable insights for educators, school boards, and other stakeholders. The book explores a diverse range of schools, highlighting their unique cultures and remarkable leaders. It also addresses universal challenges such as curriculum development, pedagogy, and meeting the social and emotional needs of students and families. The authors provide a rich and readable account of these schools, offering valuable perspectives on Jewish day school education. [Extracted from the article]

We examined the learning analytics literature over the past ten years (2011-2020). Our results show that only 33% of articles we retrieved focused on students with disabilities and 67% of articles we retrieved engaged with students with disabilities tangentially on several themes: detecting difficulties to provide early intervention, using learning analytics as tools to promote learning, addressing accessibility issues and challenges, and discussing ethics and privacy concerns. Articles that focused on students with disabilities reveal that learning analytics can be used to evaluate and optimize the design of educational technologies for students with disabilities and function as a tool to enhance learning experiences for students with various disabilities. We conclude with recommendations for future research of learning analytics community to create an inclusive platform. [ABSTRACT FROM AUTHOR]

Introduction: Online interprofessional education is a collaborative process that emphasizes both individual reflection and shared discourses. A useful analytical tool for understanding the complex dynamics of online collaborative learning is the community of inquiry (CoI) framework, which originally held that there are three types of presence in such learning: teaching, cognitive, and social. However, it was later revised to include learning presence, which is characterized by self-regulated learning. Our study aims to refine the construct of learning presence through a clearer understanding of how self- and co-regulation jointly influence learning outcomes. Methods: We surveyed 110 people involved with an online interprofessional medicaleducation curriculum at a university in Hong Kong. Path analysis was adopted to explore the relationships among 1) the three original presences of CoI; 2) learning presence (i.e., for this purpose, a combination of self-regulation and co-regulation); and 3) two learning outcomes: perceived progress and learner satisfaction. Results: The results of path analysis indicated that teaching presence had a significant indirect effect, through co-regulation, on perceived progress. In terms of direct relationships, co-regulation significantly and positively influenced both self-regulation and cognitive presence; and social presence had both positive influence on learners' satisfaction and perceived progress. Discussion: This study's findings suggest the important role of co-regulation in supporting self-regulation, especially in online collaborative-learning environments. Learners' selfregulation skills are shaped by their social interactions and regulatory activities with others. This further implies that health-professions educators and instructional designers should create learning activities that facilitate the development of co-regulatory skills, as a means of improving learning outcomes. As self-regulation is an important skill for health professions learners' lifelong learning, and because their future workplaces will be interdisciplinary in nature, it is critical to provide interactive and collaborative learning environments that will promote co-regulation and self-regulation. [ABSTRACT FROM AUTHOR]

Introduction: Students' formal networks, which are formed by a formal curriculum design, such as formally organized study groups within learning communities (LCs), may benefit students' interactions and learning. It is unclear how large-scale LCs contribute to the formation of different informal peer relationships, which refers to student self-organized out-of-class relationships. Two mechanisms can explain relationship formation in LCs. Propinquity within formal networks and homophily of students' characteristics (nationality, sex, academic performance) may promote students' peer relationships. This study explores to what extent the formation of students' informal networks was determined by their formal networks (LCs) while controlling for students' characteristics and which mechanisms play an important role. Methods: With online surveys, data were collected about five informal networks (help-seeking, collaboration, information sharing, friendship, and learn-from) from 69 first- and 51 second-bachelor year medical students (2890 relationships). Students were divided into four LCs in the formal curriculum. We compared students' five informal network structures between first- and second-year students, domestic and international students, within and between formal networks. Besides, we used Quadratic Assignment Procedure (QAP) Regression Analysis in Ucinet to investigate the associations between students' informal and formal networks (LCs) and students' characteristics. Results: Propinquity (in the same LC) plays a role since students have more informal connections within LCs than between LCs. Furthermore, it seems to play a greater role for second-year students than for first-year students. Homophily of nationality is important in informal networking since students are more likely to connect with others of similar nationalities. Conclusion: Students become more connected within the LC when they remain in the same LC for a longer period. Formal networks enhance the students' informal interactions within LCs but seem to restrict the interactions among students from other LCs. International students need support in order to integrate with domestic students in LCs. [ABSTRACT FROM AUTHOR]

The aim in this study was to identify how college students' experience of the online learning community influences their community identification and learning engagement. We used an online survey and the participants were 568 college students in the New Oriental English School's online learning community in China. The results indicated that three dimensions of community experience, namely, information experience, entertainment experience, and interactive experience, each exhibit significant positive associations with community identity, which, in turn contributes to students' learning engagement. Moreover, the finding confirmed the significant mediating effect of community identity between the three dimensions of community experience and learning engagement. The results give a deeper understanding of the role of community experience in influencing students' attitudes and behaviors in the context of an online learning community. [ABSTRACT FROM AUTHOR]