Learning Settings Research Articles

HATZFS predicts pancreatic cancer driver biomarkers by hierarchical reinforcement learning and zero-forcing set

The search for cancer biomarkers is of great significance for the early diagnosis and clinical treatment planning of pancreatic cancer. RNA transcription networks can effectively represent the interactions between RNA molecules, and the irregular perturbations in these interactions can lead to pathological states, making them commonly used for identifying cancer biomarkers. However, traditional network-based methods for identifying cancer biomarkers often rely on prior information and mutation data. Furthermore, most network controllability-based methods face the limitation of high computational complexity, preventing their integration with deep learning methods. In this paper, we propose a novel deep learning framework, called HATZFS, for identifying driver biomarkers in pancreatic cancer using RNA differential expression data. The proposed model combines hierarchical deep Q-network (HDQN), graph attention network (GAT), and zero-forcing set (ZFS). Firstly, a pancreatic cancer RNA transcription regulatory network is constructed to embed the complex relationship among Long non-coding RNA (lncRNA), microRNA (miRNA) and messenger RNA (mRNA). Then, we transform it into multiple subgraphs using graph sampling method, and apply GAT to learn the node features of the subgraphs. Finally, the HATZFS performs HDQN to select subgraphs and identify important RNA as driver nodes, and determining the controllability of the subgraphs based on ZFS. More important, the paper theoretically proves that when all subgraphs are controllable, the original graph is also controllable, and the union of the driver node sets of all subgraphs is equal to the driver node set of the original graph. To confirm the effectiveness of the proposed model, comprehensive experiments are conducted on the benchmark dataset consisting of 14 databases, comparing it with eight other algorithms. The experimental results demonstrate that the proposed model outperforms other methods. Overall, HATZFS not only provides the importance of all RNA molecules in the pancreatic cancer RNA regulatory network, but also identifies the driver node set of the network as driver biomarkers. Source code can be accessed at https://github.com/HongJieTongXue/HATZFS.

Learners Attitude and Academic Success Towards Science

Aim. COVID-19 disrupted education, leading to new learning methods. Other factors like attitude are overlooked but can aid effective learning. This paper explores two major variables: the learners’ learning attitude and academic performance in science. The study analysed students’ learning attitude, academic performance, and relationships in a blended learning setting for science. Method. The study used a mixed-methods design, with both quantitative and qualitative phases. Participants were in a blended learning program. Results. Learning can happen in a blended learning modality, with the learners’ mean grade of 87.03 (Very Satisfactory). Learners have a favourable learning attitude towards science (3.49; moderate attitude). The highest aspect of the learning attitude of learners in the blended learning modality is affective (3.72; high attitude), followed by cognitive (3.62; moderate attitude) and behavioural (3.15; moderate attitude). Furthermore, there was a positive correlation, r(78) =.48, p < 0.01, between the LLATS and SAP of the learners in blended learning. Conclusion. Learning can effectively take place in a blended learning modality. Learners exhibit a positive attitude towards science learning in a blended learning environment, with affective attitudes ranking the highest, followed by cognitive and behavioural attitudes. This can be attributed to the teachers’ efforts in providing ample opportunities for the development of both affective and cognitive aspects of attitude. However, the pandemic has presented limitations in cultivating positive behaviour attitude. Learners persevered through challenges, thanks to dedicated teachers fostering positivity and academic growth.

Graph contrastive learning as a versatile foundation for advanced scRNA-seq data analysis.

Single-cell RNA sequencing (scRNA-seq) offers unprecedented insights into transcriptome-wide gene expression at the single-cell level. Cell clustering has been long established in the analysis of scRNA-seq data to identify the groups of cells with similar expression profiles. However, cell clustering is technically challenging, as raw scRNA-seq data have various analytical issues, including high dimensionality and dropout values. Existing research has developed deep learning models, such as graph machine learning models and contrastive learning-based models, for cell clustering using scRNA-seq data and has summarized the unsupervised learning of cell clustering into a human-interpretable format. While advances in cell clustering have been profound, we are no closer to finding a simple yet effective framework for learning high-quality representations necessary for robust clustering. In this study, we propose scSimGCL, a novel framework based on the graph contrastive learning paradigm for self-supervised pretraining of graph neural networks. This framework facilitates the generation of high-quality representations crucial for cell clustering. Our scSimGCL incorporates cell-cell graph structure and contrastive learning to enhance the performance of cell clustering. Extensive experimental results on simulated and real scRNA-seq datasets suggest the superiority of the proposed scSimGCL. Moreover, clustering assignment analysis confirms the general applicability of scSimGCL, including state-of-the-art clustering algorithms. Further, ablation study and hyperparameter analysis suggest the efficacy of our network architecture with the robustness of decisions in the self-supervised learning setting. The proposed scSimGCL can serve as a robust framework for practitioners developing tools for cell clustering. The source code of scSimGCL is publicly available at https://github.com/zhangzh1328/scSimGCL.

AI fairness in practice: Paradigm, challenges, and prospects

AbstractUnderstanding and correcting algorithmic bias in artificial intelligence (AI) has become increasingly important, leading to a surge in research on AI fairness within both the AI community and broader society. Traditionally, this research operates within the constrained supervised learning paradigm, assuming the presence of class labels, independent and identically distributed (IID) data, and batch‐based learning necessitating the simultaneous availability of all training data. However, in practice, class labels may be absent due to censoring, data is often represented using non‐IID graph structures that capture connections among individual units, and data can arrive and evolve over time. These prevalent real‐world data representations limit the applicability of existing fairness literature, which typically addresses fairness in static and tabular supervised learning settings. This paper reviews recent advances in AI fairness aimed at bridging these gaps for practical deployment in real‐world scenarios. Additionally, opportunities are envisioned by highlighting the limitations and significant potential for real applications.

Online-Distance Instruction: Impact on the English Language Learning

Using a qualitative-quantitative research design, this study evaluated the influence of online- distance instruction (ODI) on the English language acquisition of college students amidst the Covid-19 pandemic at Partido State University, Camarines Sur, Philippines. The study specifically examined the effects of ODI on college students' performance in terms of: attendance, punctuality, required English language skills and competencies. Furthermore, it analyzed the students' proficiency in English through online-distance learning in the areas of: writing, grammar proficiency, oral communication, study techniques, presentation abilities, and research skills. The study also investigated the students' attitudes towards English language education in the context of online and distance learning. Ultimately, it proposed effective teaching methodologies to enhance students' English language performance in online- distance learning environments of 165 college student- participants. The findings indicate that the impact of online-distance instruction on college students' performance in terms of attendance, punctuality, and required skills and competencies is influenced by various challenges faced by the students. The students' performance in English through online-distance learning is significantly affected by the nature of online and distance learning itself. Moreover, the study revealed that students' attitudes towards English language education in online and distance learning settings are less favorable compared to traditional face-to-face instruction. To enhance students' English language performance in online-distance learning, the study recommends the implementation of three key teaching strategies: lecture method, laboratory method, and cooperative learning. These strategies are deemed beneficial for addressing the needs of student-respondents, particularly in the context of a pandemic in developing countries like the Philippines.

Advancing Indigenous primary health care research and education capacity: a qualitative study in Alberta, Canada

Background/Purpose The study explores the perspectives of family physicians in Alberta regarding Indigenous health research education within the framework of primary health care (PHC). Recognizing the key role of PHC in healthcare system, the research identified the preferences and needs of family physicians for relevant and accessible research skills education. To design effective research education, it is critical to understand the skills, content, and delivery methods desired by family physicians engaged in Indigenous health research. The Indigenous Primary Health Care and Policy Research (IPHCPR) Network is examined as a potential transformative platform aligned with Truth and Reconciliation Commission principles and Calls to Action. Methods Through purposive sampling, 12 family physicians practicing in Alberta were recruited to participate in semi-structured qualitative interviews . The study utilized a thematic content approach for data analysis. Results Four key themes emerged from the analysis: 1) Indigenous role modeling; 2) cohort-based learning; 3) relational learning; and 4) the epistemic importance of Indigenous perspectives on learning. Participants emphasized the need for mentorship, collaborative learning settings, and meaningful connections within a research network. Discussion The IPHCPR Network stands out as a pivotal platform for transformative change in healthcare research education, offering opportunities for family physicians, researchers, trainees, and communities to collaboratively improve Indigenous health outcomes. Through its inclusive practices, the network aligns with TRC principles to advance Indigenous health equity and addressing critical gaps in Indigenous primary health research and practice.

Ableism and the discourse of risk and safety in patient-facing work-integrated learning.

In many countries, such as Canada, the USA, England, and Australia, to graduate from a regulated profession such as nursing, students must complete a set of work-integrated learning (WIL) hours and demonstrate their ability to safely perform physical skills and apply knowledge in relation to professional standards. For a disabled nursing student (DNS) undergoing training in higher education institutions (HEI), securing proper accommodations to participate effectively in WIL experiences has been difficult due to concerns related to risks to self and patient safety. This study used critical discourse analysis to investigate the framing of risk and safety in association with providing DNS with accommodation in WIL. Our data were collected from an intensive codesign (group-based discussion) session with participants (n = 16), including clinicians and DNS, from four institutions (two WIL-sites and two HEI organizations). Using an iterative thematic approach based on Foucauldian framework, our analysis revealed three ways in which health professionals discursively framed risk and safety: (a) beliefs that a lack of disclosure by disabled students decreases patient safety and indicates poor self-reflection, (b) concerns that accommodating students poses a risk to institutions offering WIL experiences and to the relationships between HEI and institutions offering WIL-sites, and (c) a framing which challenges the dominant discourse by thinking about safety and risk more expansively. Our findings suggest the first two of these framings are ableist and function to enact barriers to access for DNSs in WIL, whereas the third framing serves the goal of inclusion. Recommendations to address the current ableist discourse within the Canadian nursing context are provided, which may also be applicable across other regulated health professions.

Reciprocal cognitive and emotional interaction in STEMM university learning and teaching

University learning and teaching involves contrasting and interacting emotional experiences. Even in disciplines considered as objective as Science and Engineering, emotion plays a significant role in catalysing and sustaining learning. Although emotions are individually felt, they are socially constructed between people. This is especially relevant in group-based learning evident in much contemporary higher education. This paper applies the concept of emotion socialization to illustrate how groups of students and teachers cope and succeed in cognitively and emotionally challenging learning settings. The study is based on qualitative data collected across six STEMM university departments, from 280 students via in-situ questionnaires and from 20 teachers via group and follow-up interviews. Two key findings offer insight into processes of reciprocal influence on cognition and emotion. The first relates to ways in which students’ and teachers’ differing knowledge-related goals and relationships with knowledge influenced cognition and emotion, uncovering previously unacknowledged connections. The second relates to how students’ comparisons of progress towards academic goals with peers contributed considerably to their emotional experiences in cognitively and emotionally helpful and unhelpful ways. Practical implications are presented, including possibilities for capitalising on existing emotion socialization processes and enhancing how they influence cognition and emotional well-being.

No signs of interference between learned attentional sets

No signs of interference between learned attentional sets

Designing and Testing a Contextual Factors-Based Teaching and Learning Model for Blended Mathematics Instruction

The study employs design-based research (DBR) to design the mathematics teaching and learning activity model (M-TLAM) for blended learning instruction in a Tanzanian higher education context. This model utilises contextual factors, including ICT tool usage, collaborative learning with metacognitive activities, and local culture, to determine how optimising these factors boosts math students’ motivation and achievement. Two DBR phases were conducted, of which Phase 1 lasted for two weeks and Phase 2 an entire semester. To evaluate M-TLAM in real-life learning settings, experiments were conducted with 225 first-year undergraduate students and seven lecturers at the College of Business Education in Tanzania. Experimental data were collected from pre-and post-tests, interviews, and questionnaires administered to students and lecturers. The study examined the perceptions and motivation of the participating lecturers towards using the M-TLAM in mathematics education and the factors influencing student motivation, academic achievement, and experience towards business mathematics courses. The evaluation results are promising and show that the M-TLAM implementation can potentially improve students’ motivation and academic achievement. In addition, the pedagogical experiences of students were primarily positive, and student’s attitudes towards the business mathematics course through M-TLAM were more favourable than those of students who studied the course via traditional methods. Cultural, technological, and instructional Factors contribute to students’ improved motivation. Lectures demonstrated a cheerful disposition towards the potential of the M-TLAM for enhancing teaching and learning activities. The contributions of this study are highlighted through the implementation of the M-TLAM for blended learning and through the design principles and guidelines provided towards its effective implementation.

Multitype view of knowledge contrastive learning for recommendation

Graph Neural Networks (GNNs) are playing an increasingly vital role in the field of recommender systems. To improve knowledge perception within GNNs, contrastive learning has been applied and has proven to be highly effective. GNNs have the ability to aggregate diverse knowledge and integrate topologies, while contrastive learning seeks supervisory signals from the model data. The combination of GNNs and contrastive learning can improve recommendations. However, thoughtless or incomplete contrastive learning settings limit the effectiveness of GNNs-based recommender systems in learning knowledge from knowledge and interaction graphs. To better exploit the valuable information within knowledge graphs, we propose a novel multitype view of knowledge contrastive learning for recommendations (MVKC) model.The MVKC model generates hierarchical views and augmented views in two modules, performing cross-hierarchical-view and cross-augmented-view contrastive learning and mining graph features in a self-supervised manner. The hierarchical views consist of global and local parts at multiple levels, while the augmented views are fused from the augmented knowledge graph and augmented interaction graph in our augmented processing. These features allow the MVKC model to alleviate the sparsity of user–item interaction graphs, suppress knowledge graph noise, and filter long-tail entities, which has been proven extremely important for a recommendation. The MVKC model also has strong anti-interference ability and robustness, which is crucial for a well-established model. Our experiments with three public datasets demonstrate that the MVKC model outperforms current state-of-the-art methods.

Anticipated variability increases generalization of predictive learning

We show that learners generalized more broadly around the learned stimulus when they expected more variability between the learning set and the generalization set, as well as within the generalization set. Experiments 1 and 3 used a predictive learning task and demonstrated border perceptual generalization both when expected variability was manipulated explicitly via instructions (Experiment 1), and implicitly by increasing temporal distance to the anticipated application of learning (Experiment 3). Experiment 2 showed that expecting to apply learning in the more distant future increases expected variability in the generalization set. We explain the relation between expected variability and generalization as an accuracy-applicability trade-off: when learners anticipate more variable generalization targets, they “cast a wider net” during learning, by attributing the outcome to a broader range of stimuli. The use of more abstract, broader categories when anticipating a more distant future application aligns with Construal Level Theory of psychological distance.

Learning to learn for few-shot continual active learning

Continual learning strives to ensure stability in solving previously seen tasks while demonstrating plasticity in a novel domain. Recent advances in continual learning are mostly confined to a supervised learning setting, especially in NLP domain. In this work, we consider a few-shot continual active learning setting where labeled data are inadequate, and unlabeled data are abundant but with a limited annotation budget. We exploit meta-learning and propose a method, called Meta-Continual Active Learning. This method sequentially queries the most informative examples from a pool of unlabeled data for annotation to enhance task-specific performance and tackles continual learning problems through a meta-objective. Specifically, we employ meta-learning and experience replay to address inter-task confusion and catastrophic forgetting. We further incorporate textual augmentations to avoid memory over-fitting caused by experience replay and sample queries, thereby ensuring generalization. We conduct extensive experiments on benchmark text classification datasets from diverse domains to validate the feasibility and effectiveness of meta-continual active learning. We also analyze the impact of different active learning strategies on various meta continual learning models. The experimental results demonstrate that introducing randomness into sample selection is the best default strategy for maintaining generalization in meta-continual learning framework.

Using Google Sites to Improve Students' Learning Outcomes in Writing Ability: A Case on Teaching Language Education

The ultimate purpose of this research is to investigate how far Google Sites can develop students' learning outcomes in writing ability. The need for English becomes one of the reasons why the requirement of universities in public consisting of speaking English. This research was conducted as a case study, with the objective of investigating the integration of Google Sites into language teaching strategies, specifically the design and execution of creating web-based instructional resources in Google Sites, and the benefits it provides in inspiring students to enhance their authentic writing skills. This research included 31 second-semester students, all of whom were enrolled in one of the universities in Indonesia. Among these students, 19 were female (61.29%) and 12 were male (38.71%), and they were all enrolled in the Department of English Education. These individuals were chosen based on their prior experience in language education, as they had received instruction in the subject. Based on the outcomes of the teaching treatment that utilized Google Sites to enhance students' writing skills, it can be inferred that Google Sites serves as an effective media blog for improving students' writing abilities. These results show the effectiveness of using Google Sites in both face-to-face and non-face-to-face learning settings for teaching English to students with various text materials and categories, as determined by a p-value of less than 0.01. Therefore, Google Sites can be considered an effective tool for teaching English.

CELEBRATE Feeding: A Responsive Approach to Food and Feeding in Early Learning Settings.

Early learning and child care (ELCC) settings in Canada follow nutrition standards that outline food provisions, with many also encouraging responsive feeding practices that help to create a supportive environment for children. Caregivers who lack confidence in children's ability to regulate their own intake, or those who feel stressed about mealtime, may unknowingly engage in less responsive feeding practices. The CELEBRATE Feeding Approach is a flexible framework, driven by behaviour change theory, that builds on previous definitions and concepts of responsive feeding in ELCC environments. Through this approach, there is an intentional focus on supporting early childhood educators to implement feeding practices that are more responsive. The approach incorporates 13 target educator behaviours related to the three overlapping categories of CELEBRATE language, CELEBRATE Mealtime, and CELEBRATE Play. These practices recognize and support the development of a child's sense of autonomy, confidence, and self-regulation not only at mealtimes but also through play-based exploration and language that is used throughout the day around food and feeding. The goal is that children will be open to a wide variety of food, develop their self-regulation skills, and build the foundation for a positive relationship with food throughout their lifetime.

Enjoyment in foreign language learning: A systematic review

In recent years, research on enjoyment in foreign language (FL) learning has flourished. To help illuminate the existing scope of inquiry and guide future research, this paper presents a systematic review of 118 empirical studies on FL learning enjoyment published between 2014 and 2023. Each study was coded according to its research context, methodological features, and research focus. The results indicate (1) a heavy focus on adult English as a foreign language (EFL) learners whose first languages are Chinese or Persian within traditional classroom learning settings; (2) a strong preference for quantitative methods; and (3) a prominent focus on enjoyment's antecedents and effects. Drawing upon these findings, we recommend that future research (1) addresses the experiences of language learners from diverse demographic backgrounds in a wider variety of learning settings; (2) applies multimodal methods to thoroughly assess the experience of enjoyment from both objective and subjective perspectives; and (3) explores the nature of enjoyable teacher-student or student-student socio-emotional interaction in greater depth.

An informal and transformative learning environment: The case of ODTÜ Bostan community garden in Turkey

This study explores the role of ODTÜ Bostan (METU Garden), a community garden at the Middle East Technical University in Turkey, in promoting sustainability education, ecology, and food consumption practices. For ten years, ODTÜ Bostan has been a collaborative space where participants—including students, academic and non-academic staff, and their families—grow and share food using ecological agricultural methods. This research examines how this community garden creates a setting for alternative and interdisciplinary learning, supporting practical engagement, critical thinking, and community interaction for community participants. Drawing on the Education for Sustainable Development and transformative learning frameworks, this qualitative study employs autoethnography, thematic analysis of eight semi-structured interviews, and content analysis of archival data. The findings indicate that ODTÜ Bostan not only enhances participants' sustainability competencies but also serves as an educational laboratory, demonstrating the potential of community gardens to address global challenges such as climate change and biodiversity loss. The study highlights the importance of integrating interdisciplinary and learner-centered approaches in higher education leading to collective action for societal change toward sustainability.

Twenty years of action learning in the journal Action Learning: Research and Practice

ABSTRACT This paper provides a thematic analysis of research papers published in Action Learning: Research and Practice over the past 20 years. The analysis concentrates on the different contexts in which action learning has taken place and on the variations of action learning that have been presented in different papers. Contexts discussed in this paper are different international settings for action learning, and the use of action learning in healthcare, for social action, for small and medium-sized enterprises, and in higher education. Variations and developments of action learning are critical action learning, virtual action learning, and action learning research.

Intentional learning establishes multiple attentional sets that simultaneously guide attention.

One of the key human cognitive capabilities is to extract regularities from the environment to guide behavior. An attentional set for a target feature can be established through statistical learning of probabilistic target associations; however, whether an array of attentional sets of predictive target features can be established during intentional learning, and how they might guide attention, is not known yet. To address these questions, we had human observers perform a visual search task where we instructed them to try to use color to find their target shape. We structured the task with a fine-grained statistical regularity such that the target shapes appeared in different colors with five unique probabilities (i.e., 33%, 26%, 19%, 12%, and 5%) while we recorded their electroencephalogram. Observers rapidly learned these regularities, evidenced by being faster to report targets that appeared in higher probability colors. These effects were not due to unequal sample sizes or simple feature priming. More importantly, equivalent speeding across a set of high-probability colors suggests that the brain was driving attention to multiple targets simultaneously. Our electrophysiological results showed larger amplitude N2 posterior contralateral component, indexing perceptual attention, and late positive complex (LPC) component, indexing postperceptual processes, for targets paired with high-probability colors. These electrophysiological data suggest that the learned attentional sets change both perceptual selection and how postperceptual decisions are made. In sum, we show that multiple attentional sets can be established during intentional learning that accompanies general task acquisition and that these attentional sets can simultaneously guide attention by enhancing both perceptual attention and postperceptual processes. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Speaking accuracy and fluency among EFL learners: The role of creative thinking, emotional intelligence, and academic enthusiasm

This study examines the associations between speaking accuracy and fluency among English as a Foreign Language (EFL) learners and emotional intelligence (EI), creative thinking (CT), and academic enthusiasm (AE). Quantitative data was used in a thorough analysis that was carried out utilizing a cross-sectional design. Three hundred and twenty EFL learners were chosen as a sample using a multi-stage cluster sampling technique. The creativity questionnaire by Abedi, the EI questionnaire by Schutte et al. and the AE questionnaire by Fredricks et al. were all completed by the participants. Quantitative data analysis was conducted using SPSS 25.0 and AMOS 24.0. The results showed a strong positive association between better speaking accuracy and fluency and greater levels of CT, EI, and AE. More precisely, learners with greater EI were able to control their emotions better, which improved their ability to speak clearly and fluently. Speaking performance was improved by those with higher CT because they showed improved problem-solving abilities and linguistic originality. High AE was also associated with better language practice and competence due to higher desire and involvement. These findings imply that encouraging CT, EI, and AE in language learning settings can greatly improve learners’ speaking ability. To better understand the impact of other elements on speaking proficiency, future research should look into motivation, self-esteem, and learning methodologies.