Learning Progress Research Articles

Quintuple-based Representation Learning for Bipartite Heterogeneous Networks

Recent years have seen rapid progress in network representation learning, which removes the need for burdensome feature engineering and facilitates downstream network-based tasks. In reality, networks often exhibit heterogeneity, which means there may exist multiple types of nodes and interactions. Heterogeneous networks raise new challenges to representation learning, as the awareness of node and edge types is required. In this article, we study a basic building block of general heterogeneous networks, the heterogeneous networks with two types of nodes. Many problems can be solved by decomposing general heterogeneous networks into multiple bipartite ones. Recently, to overcome the demerits of non-metric measures used in the embedding space, metric learning-based approaches have been leveraged to tackle heterogeneous network representation learning. These approaches first generate triplets of samples, in which an anchor node, a positive counterpart, and a negative one co-exist, and then try to pull closer positive samples and push away negative ones. However, when dealing with heterogeneous networks, even the simplest two-typed ones, triplets cannot simultaneously involve both positive and negative samples from different parts of networks. To address this incompatibility of triplet-based metric learning, in this article, we propose a novel quintuple-based method for learning node representations in bipartite heterogeneous networks. Specifically, we generate quintuples that contain positive and negative samples from two different parts of networks. And we formulate two learning objectives that accommodate quintuple-based learning samples, a proximity-based loss that models the relations in quintuples by sigmoid probabilities and an angular loss that more robustly maintains similarity structures. In addition, we also parameterize feature learning by using one-dimensional convolution operators around nodes’ neighborhoods. Compared with eight methods, extensive experiments on two downstream tasks manifest the effectiveness of our approach.

Nature of science and domain-specificity: investigating the coverage of nature of science in physics, chemistry and biology curricula across grade levels

ABSTRACT The paper investigates the coverage of nature of science (NOS) in different science subject curricula to understand how they emphasise various aspects of NOS across grade levels in secondary science education. Although various science curriculum analyses exist in the literature, there is a paucity of investigations that have combined the elements of domain-specificity and learning progression in relation to NOS curriculum content. Furthermore, a holistic and inclusive account of NOS is used to examine the cognitive, epistemic, social and institutional aspects of NOS. Physics, chemistry and biology high school curricula from Turkey are investigated. Each subject curriculum is analysed and compared using an analytical framework guided by the Family Resemblance Approach. Mixed methods approaches are used to summarise the quantitative and qualitative results. Findings suggest an overemphasis on the cognitive-epistemic aspects of NOS, considerable variation across science subject areas and non-uniform progression of NOS themes across grade levels. Analysis of curricula incorporating the domain-specific and learning progression elements of NOS is likely to enhance understanding of how to design robust curricula in different science subjects and to improve the teaching and learning of NOS. Implications for curriculum reform are discussed.

Cross‐attention neural network for land cover change detection with remote sensing images

AbstractLand cover change detection (LCCD) with remote sensing images (RSIs) is important for observing the land cover change of the Earth's surface. Considering the insufficient performance of the traditional self‐attention mechanism used in a neural network to smoothen the noise of LCCD with RSIs, in this study a novel cross‐attention neural network (CANN) was proposed for improving the performance of LCCD with RSIs. In the proposed CANN, a cross‐attention mechanism was achieved by employing another temporal image to enhance attention performance and improve detection accuracies. First, a feature difference module was embedded in the backbone of the proposed CANN to generate a change magnitude image and guide the learning progress. A self‐attention module based on the cross‐attention mechanism was then proposed and embedded in the encoder of the proposed network to make the network pay attention to the changed area. Finally, the encoded features were decoded to obtain binary change detection with the ArgMax function. Compared with five methods, the experimental results based on six pairs of real RSIs well demonstrated the feasibility and superiority of the proposed network for achieving LCCD with RSIs. For example, the improvement for overall accuracy for the six pairs of real RSIs improved by our proposed approach is about 0.72–2.56%.

A Study of Learning Progression in Science of Class IX Students of Balangir District, Odisha

A Study of Learning Progression in Science of Class IX Students of Balangir District, Odisha

“Learning from Each Other”: A Peer Education as an Alternative Teaching Option

This article presents a model of teaching as a new entry in the relevant bibliography, under the name “Learning from Each Other” (LEO) Teaching Model. The main purpose of LEO is the cognitive, emotional and social interaction and cooperation between learners of different ages and genders along with the mutual transmission of individual opinions, knowledge, attitudes, experiences, abilities and skills of learners, in the context of simultaneously taking on the role of the “tutor” and the “learner” as well. At the same time, the role of the teacher during the teaching is being revised by becoming not only a “teacher” but also a “learner.” Thus, for the teacher there are options about new researches and reflections regarding teaching as well as the progress and value of learning.

Simulation of E-learning in English personalized learning recommendation system based on Markov chain algorithm and adaptive learning algorithm

The research topic of this article is the English personalized learning recommendation module based on Markov chain algorithm and adaptive learning algorithm. Personalized learning recommendation systems have been widely applied in the field of education. Firstly, this article uses Markov chain algorithm to analyze user learning behavior data, abstract user learning behavior as a process of state transition, predict their future behavior based on their past behavior, better understand their learning needs and interests, and provide more personalized learning recommendations for them. Subsequently, combined with adaptive learning algorithms, recommendation strategies and content are dynamically adjusted based on the user's learning goals and level, providing more accurate and effective learning recommendations according to their personalized needs and learning progress. By comparing with traditional recommendation systems, evaluate the accuracy and personalization of the recommendation module to better meet the learning needs of users.

Control of Qubit Dynamics Using Reinforcement Learning

The progress in machine learning during the last decade has had a considerable impact on many areas of science and technology, including quantum technology. This work explores the application of reinforcement learning (RL) methods to the quantum control problem of state transfer in a single qubit. The goal is to create an RL agent that learns an optimal policy and thus discovers optimal pulses to control the qubit. The most crucial step is to mathematically formulate the problem of interest as a Markov decision process (MDP). This enables the use of RL algorithms to solve the quantum control problem. Deep learning and the use of deep neural networks provide the freedom to employ continuous action and state spaces, offering the expressivity and generalization of the process. This flexibility helps to formulate the quantum state transfer problem as an MDP in several different ways. All the developed methodologies are applied to the fundamental problem of population inversion in a qubit. In most cases, the derived optimal pulses achieve fidelity equal to or higher than 0.9999, as required by quantum computing applications. The present methods can be easily extended to quantum systems with more energy levels and may be used for the efficient control of collections of qubits and to counteract the effect of noise, which are important topics for quantum sensing applications.

A Hybrid Deep Learning Architecture for Apple Foliar Disease Detection

Incorrectly diagnosing plant diseases can lead to various undesirable outcomes. This includes the potential for the misuse of unsuitable herbicides, resulting in harm to both plants and the environment. Examining plant diseases visually is a complex and challenging procedure that demands considerable time and resources. Moreover, it necessitates keen observational skills from agronomists and plant pathologists. Precise identification of plant diseases is crucial to enhance crop yields, ultimately guaranteeing the quality and quantity of production. The latest progress in deep learning (DL) models has demonstrated encouraging outcomes in the identification and classification of plant diseases. In the context of this study, we introduce a novel hybrid deep learning architecture named “CTPlantNet”. This architecture employs convolutional neural network (CNN) models and a vision transformer model to efficiently classify plant foliar diseases, contributing to the advancement of disease classification methods in the field of plant pathology research. This study utilizes two open-access datasets. The first one is the Plant Pathology 2020-FGVC-7 dataset, comprising a total of 3526 images depicting apple leaves and divided into four distinct classes: healthy, scab, rust, and multiple. The second dataset is Plant Pathology 2021-FGVC-8, containing 18,632 images classified into six categories: healthy, scab, rust, powdery mildew, frog eye spot, and complex. The proposed architecture demonstrated remarkable performance across both datasets, outperforming state-of-the-art models with an accuracy (ACC) of 98.28% for Plant Pathology 2020-FGVC-7 and 95.96% for Plant Pathology 2021-FGVC-8.

Structures of neural network effective theories

We develop a diagrammatic approach to effective field theories (EFTs) corresponding to deep neural networks at initialization, which dramatically simplifies computations of finite-width corrections to neuron statistics. The structures of EFT calculations make it transparent that a single condition governs criticality of all connected correlators of neuron preactivations. Understanding of such EFTs may facilitate progress in both deep learning and field theory simulations. Published by the American Physical Society 2024

Integrating Artificial Intelligence (AI) Into Adult Education

This conceptual article provides a comprehensive overview of the current status of Artificial Intelligence (AI) integration and its influence on adult education. It discusses generative AI technologies and their potential applications in adult education settings, examines the opportunities and ethical challenges associated with integrating AI, and provides insights into emerging trends. The article consists of five sections. The introduction provides a rationale as to why AI should be integrated into adult education. Second, it describes evolving AI technologies such as Large Language Models (LLM) for personalized learning, Machine Learning Algorithms for adaptive learning systems, Virtual Reality (VR) and Augmented Reality (AR) for immersive learning experiences, Chatbots and virtual assistants for learner support and guidance, and Data Learning Analytics (DLA) for tracking learner progress and performance into adult education. Section three explores the ethical implications of AI in adult education, including academic honesty and integrity, data privacy, and algorithmic bias. In section four, emerging trends and future directions are discussed. The final section considers policy implications and makes recommendations for adult educators working to develop AI-enriched adult education.

Modeling and simulation of a double DQN algorithm for dynamic obstacle avoidance in autonomous vehicle navigation

Developing an autonomous vehicle in safe navigated open environments through self-learning is a formidable challenge. So, in this paper, an autonomous navigation implementation for urban environments with dynamic obstacles using a double DQN algorithm is proposed. Furthermore, the vehicle is trained in varied training environments, explored neural network architectures, and fine-tuned action and reward functions to optimize its performance. In conclusion, the vehicle performance is steadily improved with each iteration of training, as evident from the upward trend in the average reward graph. These graphs depict the vehicle's learning progress in the scenario where it is trained to avoid dynamic obstacles. Finally, the proposed algorithm shows better performance relative to the conventional algorithms. The entire work is modelled and simulated in CARLA software.

Learning outcome evaluation in manual assembly

Mass customization and shorter product life cycles are causing ever more variants in production, especially in manual assembly. At the same time, more diverse personnel structures are emerging due to demographic change and labor shortages. This is causing different challenges to production managers, e.g., competence gaps. To meet these challenges, learning in manual assembly becomes increasingly important. The design of the learning process can only be improved by checking whether the processes fulfill their purpose. Various learning evaluation measures are described in general vocational education and competence development, but it is hard to select the right one for the learning process. This paper shows a procedure, how learning evaluation measures can be selected, and how they can measure learning progress. For this, a test person study was conducted to compare different learning evaluation measures and show their usability and advantages in manual assembly. The results support making learning in assembly easier to apply and controllable. In the long term, feeding back the results improves the learning process design.

Connecting the threads: the role of multiplicative thinking in algebraic, geometrical, and statistical reasoning

ABSTRACT Making connections within and between different aspects of mathematics is recognised as fundamental to learning mathematics with understanding. However, exactly what these connections are and how they serve the goal of learning mathematics is rarely made explicit in curriculum documents with the result that mathematics tends to be presented as a set of discrete, disconnected topics. Interest in establishing a more coherent approach to the teaching and learning of school mathematics has led to a focus on big ideas. That is, networks of related concepts, skills and ways of thinking that facilitate learning mathematics with understanding. Research on learning progressions has helped identify what these big ideas are and how they serve to build connections within and between different aspects of mathematics. This paper draws on research that provides an evidenced-based learning progression for multiplicative reasoning to illustrate the connective role of multiplicative thinking in the development of algebraic, geometrical, and statistical reasoning.

Learning disruptions during the COVID-19 pandemic: Evidence from household surveys in Southeast Asia

We study children’s access to remote learning when schools were closed during the COVID-19 pandemic and their parents’ perceptions about learning progress in seven Southeast Asian countries. This is the first regional analysis to systematically document students’ access to remote learning based on survey data and to investigate how school closures and remote learning access affected children’s learning progress. The results are based on survey responses from 2200 households. We find that 79% of the respondents felt that their children’s learning progress was slower during school closures than it would have been with in-person schooling. Slightly less than half of all children experienced very little or no learning progress. Three characteristics were strongly correlated with learning progress. First, boys were more likely than girls to experience very little or no progress. Second, children from households in the top 30% of the income distribution were more likely to progress at the same rate as in-person classes, compared to children from lower income households. Third, comparing the different remote learning modes, internet-based learning or multiple learning modes provided children with a better chance of maintaining learning progress than other single modes.

The Development and Implementation of a Competency-Focused Curriculum in Health and Exercise Sciences

Historically, undergraduate programs in Health and Exercise Sciences (i.e. Kinesiology) are rooted in common areas such as physiology, anatomy, biomechanics, motor control, psychology, nutrition, and rehabilitation. A typical four-year program offers a general education across these areas often delivering a breadth of knowledge in many areas, albeit often lacking in transferable technical skill that relates to industry. With many graduates now pursuing advanced training in professional programs there is a need to develop more competency-focused curricula to better prepare students for entry into professional fields. Further, overburdened healthcare systems worldwide combined with an ageing population find themselves with a gap in service delivery, given the growing body of evidence supporting the benefits of high-quality exercise training in the prevention, treatment and management of chronic conditions. In 2018, following feedback from provincial ministries, the University Senate, faculty, students, and community partners, the School of Health and Exercise Sciences at the University of British Columbia Okanagan underwent a strategic review. This review led to the development of a professionally focused, competency-based program, now known as the Bachelor of Health and Exercise Sciences. Within this program are three concentrations geared towards developing knowledge and skills in differing areas dependent on student interest and desired career path: Kinesiology and Allied Health, Health Behaviour Change, and Clinical Exercise Physiology. The development of the revised curriculum involved crucial components such as defining new program and concentration level learning outcomes, competency profiles, definitions, progressions, and practical assessment strategies. Additionally, each concentration integrates a community practicum designed to enhance mastery of competencies within a professional setting. Establishing these practicums necessitates ongoing collaborative efforts with both existing and new community partners, the formulation of standards, and the implementation of a process to ensure students possess a baseline skill level before engaging with the community. A substantial challenge in the revised curriculum is the instruction of technical skills in a lab environment, as well as effective practical skill assessment that meets industry standards, for upwards of 200 students. To this end a great deal of planning was required from a resources standpoint, onboarding of new faculty with industry experience, as well as alignment with pre-existing certifying bodies including the American College of Sports Medicine, Exercise & Sport Science Australia, and the Canadian Society for Exercise Physiology. Looking to the future we anticipate both challenges and opportunities as we continue to redefine professional competencies and remap them throughout the curriculum to enhance skill progression and ensure our graduates are fit for purpose. Further, we are excited to continue to work alongside provincial entities and existing certifying bodies as we look toward the professionalization of the field. Lastly, but up utmost importance, is the dissemination of our process to other organizations and institutions in hopes to build momentum in the quickly evolving academic and professional landscape of Health and Exercise Sciences. None. This is the full abstract presented at the American Physiology Summit 2024 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.

Levels of teacher performance in formative assessment in multigrade and single-grade classrooms

Formative assessment is an evaluative practice developed in the classroom for the improvement of learning using evidence on student progression. The objective of this research is to compare sample groups from multigrade and single-grade classrooms on the theme of formative assessment based on the students' opinion of the teacher's performance. The method used was a comparative quantitative method. The sample type is a probability sample of 683 students from 5th to 8th grade from urban and rural schools in the commune of Longaví, located in the Maule Region of Chile. A validated Likert scale questionnaire with a high level of reliability (α = 0.93) was used. The results of the research showed that, in the six dimensions, the best teacher performance concerning formative assessment is found in multi-grade schools and not in single-grade schools. This can be explained on the basis of several reasons, among them the level of adaptability that teachers have in this type of classroom, the heterogeneous characteristics of the classroom (different ages and learning goals) and the need for teachers to monitor the learning progression of students with different classroom characteristics.

More than 20 procedures are necessary to learn small bowel capsule endoscopy: Learning curve pilot study of 535 trainee cases.

Background and study aims The number of procedures needed to acquire a sufficient level of skills to perform an unassisted evaluation of small bowel capsule endoscopy (SBCE) is unknown. We aimed to establish learning curves, diagnostic accuracy, and the number of procedures needed for reviewing small bowel capsule endoscopies unassisted. Methods An expert panel developed a 1-day course including lessons (examination, anatomy, and pathology) and hands-on training. After completing the course, participants received 50 cases in a randomized sequence. An interactive questionnaire about landmarks, findings, and diagnosis followed each case. After submitting the questionnaire, participants received feedback. Data are presented using CUSUM (cumulative sum control chart) learning curves and sensitivity/specificity analyses compared with expert opinions. Results We included 22 gastroenterologists from 11 different Danish hospitals. A total of 535 cases were reviewed (mean: 28; range: 11-50). CUSUM plots demonstrated learning progression for diagnosis and findings during the course, but none of the participants reached a learning plateau with sufficient competencies. The sensitivity for all findings was 65% (95% confidence interval [CI] 0.51-0.82) for the first 20 procedures and 67% (95% CI 0.58-0.73) from case 21 until completion or dropout. The specificity was 63% (95% CI 0.52-0.74) for the first 20 procedures and 57% (95% CI 0.37-0.77) for the rest. Conclusions Our data indicate that learning SBCE may be more difficult than previously recognized due to low discriminative abilities after 20 cases except for the identification of CD. This indicates that 20 SBCE cases may not be sufficient to achieve competency for reviewing SBCE without supervision.

Influencing Factors of Chinese Proficiency Test Learning Application Under the Framework of Seamless Learning

With the development of modern information technology and network society, the era of digital transformation has arrived, many educational resources have been developed and utilized, and the problem of space-time separation of global Chinese learners is gradually being solved. The use of mobile application learning can achieve convenient and fast learning, reduce learning costs, and flexibly adjust learning progress. After experiencing COVID-19, the demand for human-computer interaction in mobile learning is even more necessary. Therefore, to make applications achieve seamless docking with the needs of learners in the design of learning, the research investigates and analyzes the five dimensions of time, space, mode, subject, and object. The research tests the reliability and validity of the questionnaire, removes the relevant items and conducts non-parametric tests, then puts forward corresponding strategies according to the results of the questionnaire. It aims to enhance the efficiency of Chinese learners in HSK (Hanyu Shuiping Kaoshi, Chinese Proficiency Test) learning and promote Chinese learning and the spread of the Chinese language and culture.

The Structure of Open and Distance English Language Teaching Models: A Follow Up Study 2018 – 2023

In recent years, the widespread adoption of Open and Distance Learning (ODL) has revolutionized education, especially English language teaching. ODL has democratized access to educational resources and enables interaction between students and instructors across geographical boundaries. Moreover, technology integration has emerged as a dominant theme, with a focus on web-based, computeraided, and mobile-learning systems. However, the study advocates for innovative approaches, such as mobile and game-based learning, aligning with contemporary trends. The adoption of various evaluation methods shows the recognition of individual differences and underscores the need for a comprehensive view of learner progress. This research employs a systematic approach to review literature and investigate the common structures of ODL models for teaching the English language, focusing on articles published between 2018 and 2023. This study categorizes studies based on five macro dimensions of ODL —management, communication, learning, technology, and evaluation—in order to identify common themes and patterns within the literature. In contrast to a previous study spanning 1990 to 2017, this investigation uncovers the evolving landscape of distance education, underscoring the need for continued research and improvement. This research emphasizes the lack of a comprehensive management model for EFL education delivered through ODL, and suggests the need for further investigation in this domain. Moreover, the findings emphasize the significance of leveraging diverse communication models to enhance learner engagement. A shift towards autonomous learning models is observed, demanding further investigation into strategies that support effective self-directed learning experiences. In conclusion, this comprehensive analysis provides a roadmap for advancing ODL models in English-language education. To enhance the quality and accessibility of English language instruction in ODL, educators and researchers can concentrate on addressing management gaps, optimizing communication strategies, promoting autonomous learning, exploring advanced technologies, and refining evaluation approaches. This research offers a rich repository of knowledge for educators, researchers, and academic institutions that wish to deepen their comprehension and practical application of ODL approaches in the domain of English language instruction

Advanced Recognizing and Categorizing System for Alzheimer's and Parkinson's Syndromes using CNN Algorithm and Clinical Data

Parkinson’s Disease (PD) and Alzheimer’s Disease (AD) are advancing neurodegenerative diseases that impact a large number of individuals globally. This study provides a novel method to identify and categorize individuals in the early stages of PD and AD by analyzing online handwriting structures, leveraging the current progress in Machine Learning (ML) in the medical domain. The research explores multiple data augmentation strategies to address the limited training information available for classifying neuro-degenerative disorders based on behavioral information. The study used a Convolutional Neural Network (CNN) classification as a foundation model by identifying essential radiomics characteristics from the T1-weighted Magnetic Resonance Imaging (MRI) data. The research explores data augmentation strategies designed explicitly for time-series information, such as online handwriting patterns, instead of standard methods used for hardware-based detection. This approach, with more validation, might assist in distinguishing between difficult instances of AD and PD that have a comparable mild motor and non-motor symptoms.