Learning Paradigm Research Articles

A Metric-Based Detection System for Large Language Model Texts

More efforts are being put into improving the capabilities of Large Language Models (LLM) than into dealing with their implications. Current LLMs are able to generate high-quality texts seemingly indistinguishable from those written by human experts. While offering great potential, such breakthroughs also pose new challenges for safe and ethical uses of LLMs in education, science, and a multitude of other areas. Thus, majority of current approaches in LLM text detection are either computationally expensive or need access to the LLMs’ internal computations, both of which hinder their public accessibility. With such motivation, this article presents a novel metric learning paradigm for detection of LLM-generated texts that is able to balance computational costs, accessibility, and performances. Specifically, the detection is based on learning a similarity function between a given text and an equivalent example generated by LLMs that outputs high values for LLM-LLM text pairs and low values for LLM-human text pairs. In terms of architecture, the detection framework includes a pre-trained language model for the text embedding task and a newly designed deep metric model. The metric component can be trained on triplets or pairs of same-context instances to signify the distances between human and LLM texts while reducing that among LLM texts. Next, we develop five datasets totaling more than 95,000 contexts and triplets of responses in which one is from humans and two are from GPT-3.5 TURBO or GPT-4 TURBO for benchmarking. Experiment studies show that our best architectures maintain F1 scores between 0.87 and 0.95 across the tested corpora in multiple experiment settings. The metric framework also demands significantly less time in training and inference compared to RoBERTa, LLaMA 3, Mistral v0.3, and Ghostbuster, while keeping 90% to 150% performance of the best benchmark.

Alternating-update-strategy based Graph Autoencoder for graph neural network

Abstract Self-supervised learning (SSL) has become a promising and popular learning paradigm for graph data, offering the advantage of capturing informative knowledge without reliance on manual labels. As a representative class of generative graph SSL models, existing graph autoencoders (GAE) excel in link prediction tasks and are steadily improving in node classification tasks. However, GAE is essentially based on the Information Maximization (InfoMax) principle, always captures much redundant information. In this paper, we propose an Alternating-update-strategy based Graph Autoencoder, including alternating update module (AUM) and GAE. For AUM, we design an Alternating-update-strategy to generate a new graph with reduced redundancy, in order to reduce the amount of redundant information that the encoder may capture. For GAE, we feed it the new graph and employ a re-mask decoding strategy to generate node representations. Our model is evaluated on five common real-world datasets for the node classification task, and the experimental results demonstrate its superiority. Meanwhile, our model has also achieved excellent results in specific e-commerce warehousing application scenarios.

Computer Ability Anxiety, Self-Efficacy Beliefs in Statistical Studies, Student Attitudes towards Course Statistics: An Empirical Study of Undergraduate Students in the Mathematics Study Program

In the 21st century, the learning process continues to develop and undergo a paradigm shift. One of the characteristics of the 21st century is the presence of aspects of information, computing, automation, and communication that influence the learning paradigm. In learning statistics, it is not only cognitive skills that support student learning success but also non-cognitive skills. This study aims to determine the effect of student computer ability anxiety and self-efficacy beliefs on student attitudes toward statistical courses. The sample of this study was 203 students of the Mathematics Study Program of Pamulang University who were selected using a simple random sampling technique. Data collection techniques in this study used instruments with a Likert scale. The effect on the research variables was analyzed using structural equation modeling to determine the effect. From the results of the analysis, it is obtained that there is a direct influence of computer ability anxiety on self-efficacy beliefs in learning statistics, there is a direct influence of computer ability anxiety on attitudes towards statistics courses, there is a direct influence of self-efficacy beliefs in learning statistics on attitudes towards statistics courses, and there is an influence of computer ability anxiety in learning statistics on attitudes towards statistics courses through self-efficacy beliefs.

AI-ML based HealthCare Chatbot

In the realm of modern healthcare, the exigency for instantaneous, accurate, and personalized medical assistance remains a paramount concern, primarily due to infrastructural deficits and delayed medical interventions, especially in remote locales. The AI-ML-powered healthcare chatbot aspires to ameliorate this predicament by offering 24/7 medical support, ranging from symptom assessment to scheduling consultations and medication reminders, thereby fortifying healthcare accessibility. The focal point of this inquiry entails a meticulous exploration of the underlying factors contributing to delayed medical responses, analysing the causative elements, and investigating the contextual framework to comprehend when and how such exigencies materialize. A comprehensive literature survey will be undertaken to discern prevailing methodologies and state-of-the-art techniques employed for the detection and prevention of health-related complications through chatbot interventions, leveraging machine learning and natural language processing paradigms. Subsequently, a comparative analysis of these techniques will be conducted to delineate their respective efficacies, performance metrics, and limitations, facilitating the selection of the most optimal strategy for deployment in real-world scenarios. Ultimately, the inquiry will culminate in a decisive conclusion that encapsulates the key findings, underscores the transformative potential of AI-ML chatbots in healthcare delivery, and delineates actionable recommendations for future enhancements and scalability . Keywords—Healthcare Chatbot, Artificial Intelligence (AI), Machine learning (ML), Disease detection, Healthcare automation, Symptom Assessment

Unique Method for Prognosis of Risk of Depressive Episodes Using Novel Measures to Model Uncertainty Under Data Privacy

The research described in this paper focuses on key aspects of learning from data concerning the symptoms of depression and how to prevent it. The computer support system designed for that purpose combines data privacy protection from various sources and uncertainty modeling, especially for incomplete data. The mentioned aspects are key to real-life medical diagnostic problems. From among the different paradigms of machine learning, a federated learning-based approach was chosen as the most suitable to take up the challenge. Importantly, computer support in medical diagnostics often requires algorithms that are appropriate for processing data expressing uncertainty and that can ensure high-quality diagnostics. To achieve this goal, a novel decision-making algorithm is used that employs interval entropy measures based on the theory of interval-valued fuzzy sets. Such an approach enables one to take into account diagnostic uncertainty, express it exactly, and interpret it easily. Furthermore, the applied classification technique offers the possibility of a straightforward explanation of the diagnosis, which is a situation required by many physicians. The presented solution combines innovative technological approaches with practical user needs, fostering the development of more effective tools in mental health prevention.

Multi-modal dataset creation for federated learning with DICOM-structured reports.

Purpose Federated training is often challenging on heterogeneous datasets due to divergent data storage options, inconsistent naming schemes, varied annotation procedures, and disparities in label quality. This is particularly evident in the emerging multi-modal learning paradigms, where dataset harmonization including a uniform data representation and filtering options are of paramount importance.Methods DICOM-structured reports enable the standardized linkage of arbitrary information beyond the imaging domain and can be used within Python deep learning pipelines with highdicom. Building on this, we developed an open platform for data integration with interactive filtering capabilities, thereby simplifying the process of creation of patient cohorts over several sites with consistent multi-modal data.Results In this study, we extend our prior work by showing its applicability to more and divergent data types, as well as streamlining datasets for federated training within an established consortium of eight university hospitals in Germany. We prove its concurrent filtering ability by creating harmonized multi-modal datasets across all locations for predicting the outcome after minimally invasive heart valve replacement. The data include imaging and waveform data (i.e., computed tomography images, electrocardiography scans) as well as annotations (i.e., calcification segmentations, and pointsets), and metadata (i.e., prostheses and pacemaker dependency).Conclusion Structured reports bridge the traditional gap between imaging systems and information systems. Utilizing the inherent DICOM reference system arbitrary data types can be queried concurrently to create meaningful cohorts for multi-centric data analysis. The graphical interface as well as example structured report templates are available at https://github.com/Cardio-AI/fl-multi-modal-dataset-creation .

Advancing IoT Security with an Innovative Machine Learning Paradigm for Botnet Attack Detection

INTRODUCTION: In contemporary society, everyday operations are greatly improved by the Internet of Things (IoT), which connects physical devices to provide digital services. IoT technology offers unified services and streamlines activities across various domains, ranging from remote monitoring to sophisticated welfare systems. However, the growing number of IoT devices presents a security concern. Many of these devices are susceptible to exploitation, leading to diverse vulnerabilities. OBJECTIVES: Resource-constrained IoT devices become prime targets for botnet attacks, manifesting in various forms and penetration methods. Despite numerous research efforts introducing multiple approaches for detecting botnet attacks in IoT, existing methods often fail to achieve satisfactory detection rates. METHODS: Additionally, these approaches struggle to comprehensively analyze the diverse communication networks within the expansive realm of IoT devices. This study proposes an innovative machine-learning framework for detecting IoT botnet threats to address these limitations. RESULTS: This conceptual framework exhibits a remarkable capability to identify a spectrum of botnet attacks, showcasing a detection accuracy of 99.5 per cent, significantly surpassing the performance of other prevalent machine-learning approaches. CONCLUSION: Through this research, we aim to enhance the security paradigm of IoT networks, ensuring robust protection against evolving botnet threats in the dynamic landscape of interconnected devices.

A novel augmented reality-based simulator for enhancing orthopedic surgical training.

Total Hip Arthroplasty (THA) is a well-established and common orthopedic surgery. Due to the complexity involved in THA, orthopedic surgeons require rigorous training. However, the current gold standard, the tutor-guided and -evaluated apprenticeship model is time-consuming, costly, and poses risks to patients. There is a pressing need for additional training resources to enhance the efficiency and safety of the training process. In this work, we present a novel Augmented Reality (AR)-based simulator designed for THA that helps enable a new self-paced training and learning paradigm without the need for instructors. The simulator reduces the need for instructors by integrating an AR guidance module and an automated performance evaluation module. Three types of AR guidance were developed: Overlay, Virtual Twin, and Sectional Views. A feasibility study was conducted with five resident surgeons and two senior surgeons to compare these guidance methods quantitatively and qualitatively. The automated performance evaluation module was assessed against manual performance evaluation using Bland-Altman analysis with limits of agreement (LoA) and Mann-Whitney U tests. The quantitative feasibility results indicate the efficacy of the developed AR guidance, characterized by mean transitional and rotational deviation errors below 3 mm and 3 degrees. Based on the qualitative results, we provide recommendations for efficient AR guidance designs. The Bland-Altman analysis results (0.22±1.32mm with LoA -2.37 to 2.81 mm for distance deviation, 0.94±2.41 degrees with LoA -3.78 to 5.66 degrees for yaw deviation, -0.34±1.30 degrees with LoA -2.90 to 2.22 degrees for pitch deviation) and p-values of Mann-Whitney U tests (0.64 for distance deviation, 0.12 for yaw deviation, 0.11 for pitch deviation) indicate no statistically significant differences between the automated and manual performance evaluation at a significance level of 0.05. This work shows the potential of AR-based simulators in introducing a novel, data-driven approach to open surgery training in orthopedics, enabling surgeons to individually assess and improve their progress.

An interpretable and explainable deep learning model for predicting hydrogen solubility in diverse chemicals

An interpretable and explainable deep learning model for predicting hydrogen solubility in diverse chemicals

Semi-supervised single-view 3D reconstruction via multi shape prior fusion strategy and self-attention

Semi-supervised single-view 3D reconstruction via multi shape prior fusion strategy and self-attention

Human-Centric Transformer for Domain Adaptive Action Recognition.

We study the domain adaptation task for action recognition, namely domain adaptive action recognition, which aims to effectively transfer action recognition power from a label-sufficient source domain to a label-free target domain. Since actions are performed by humans, it is crucial to exploit human cues in videos when recognizing actions across domains. However, existing methods are prone to losing human cues but prefer to exploit the correlation between non-human contexts and associated actions for recognition, and the contexts of interest agnostic to actions would reduce recognition performance in the target domain. To overcome this problem, we focus on uncovering human-centric action cues for domain adaptive action recognition, and our conception is to investigate two aspects of human-centric action cues, namely human cues and human-context interaction cues. Accordingly, our proposed Human-Centric Transformer (HCTransformer) develops a decoupled human-centric learning paradigm to explicitly concentrate on human-centric action cues in domain-variant video feature learning. Our HCTransformer first conducts human-aware temporal modeling by a human encoder, aiming to avoid a loss of human cues during domain-invariant video feature learning. Then, by a Transformer-like architecture, HCTransformer exploits domain-invariant and action-correlated contexts by a context encoder, and further models domain-invariant interaction between humans and action-correlated contexts. We conduct extensive experiments on three benchmarks, namely UCF-HMDB, Kinetics-NecDrone and EPIC-Kitchens-UDA, and the state-of-the-art performance demonstrates the effectiveness of our proposed HCTransformer.

Estimating oil recovery efficiency of carbonated water injection with supervised machine learning paradigms and implications for uncertainty analysis

Estimating oil recovery efficiency of carbonated water injection with supervised machine learning paradigms and implications for uncertainty analysis

Long-term exposure therapy outcome in phobia and the link with behavioral and neural indices of extinction learning.

Extinction learning is regarded as a core mechanism underlying exposure therapy. Under this assumption, studies have looked at the predictive value of the extinction learning paradigm for exposure therapy outcomes. However, predicting factors of long-term exposure therapy success have not been established. Participants with a specific phobia (SP) for spiders were included in a double-blind randomized controlled trial. Participants were randomly assigned to receive exposure therapy (n=25, 24 females) or an active control intervention, progressive muscle relaxation (PMR; n=18, 15 females). Symptom levels were measured with the Fear of Spiders questionnaire (FSQ) at baseline (T0), after the intervention (T1), and at six- (T2) and twelve (T3) months follow-up. At baseline, participants completed a three-day fMRI fear conditioning, extinction learning, and extinction recall paradigm. Indices of extinction were defined as self-reported threat expectancy and fear, and neural activation during stimulus presentations and threat omission in the ventromedial prefrontal cortex and nucleus accumbens, based on prior data. Mixed model analysis revealed that the exposure therapy group had an overall stronger decrease in phobic symptoms over time than the PMR group (β=10.95, p<.001). However, none of the indices of extinction learning were predictive for FSQ scores after exposure therapy at the longest follow-up measurement (T3). In sum, the current results show the long-term effectiveness of a single session of exposure therapy for reducing a specific fear of spiders but no baseline characteristics were identified that predicted individual differences in exposure therapy success after one year.

Kurikulum Merdeka: Harapan Baru Bagi Pendidikan Indonesia

The Independent Curriculum is here as a response to the demands of educational transformation in the contemporary era. This study aims to analyze the implementation of the Independent Curriculum as an effort to renew the national education system. Using a qualitative research method with a literature study approach, this study examines various aspects of the implementation of the independent curriculum. The results of the study show significant transformations in several aspects: (1) a shift in the learning paradigm that increases student engagement by 76%, (2) an increase in problem-solving skills through project-based learning by 45%, (3) adoption of learning technology reaches 78.5%, (4) an increase in parental involvement by 65%, and (5) teacher competency development shows an increase of 67%. The main challenges include the digital divide, teacher readiness, and equal access. Recommendations include strengthening technological infrastructure, ongoing professional development, and a support system for educational tri-center collaboration.

A survey of security threats in federated learning

Federated learning is a distributed machine learning paradigm that emerged as a solution to the need for privacy protection in artificial intelligence. Like traditional machine learning, federated learning is threatened by multiple attacks, such as backdoor attacks, Byzantine attacks, and adversarial attacks. The weaknesses are exacerbated by the inaccessibility of data in federated learning, which makes it more difficult to defend against these threats. This points to the need for further research into defensive approaches to make federated learning a real solution for distributed machine learning paradigm with securing data privacy. Our survey provides a taxonomy of these threats and defense methods, describing the general situation of this vulnerability in federated learning. We also sort out the relationship between these methods, their advantages and disadvantages, and discuss future research directions regarding the security issues of federated learning from multiple perspectives.

The calcitron: A simple neuron model that implements many learning rules via the calcium control hypothesis.

Theoretical neuroscientists and machine learning researchers have proposed a variety of learning rules to enable artificial neural networks to effectively perform both supervised and unsupervised learning tasks. It is not always clear, however, how these theoretically-derived rules relate to biological mechanisms of plasticity in the brain, or how these different rules might be mechanistically implemented in different contexts and brain regions. This study shows that the calcium control hypothesis, which relates synaptic plasticity in the brain to the calcium concentration ([Ca2+]) in dendritic spines, can produce a diverse array of learning rules. We propose a simple, perceptron-like neuron model that has four sources of [Ca2+]: local (following the activation of an excitatory synapse and confined to that synapse), heterosynaptic (resulting from the activity of other synapses), postsynaptic spike-dependent, and supervisor-dependent. We demonstrate that by modulating the plasticity thresholds and calcium influx from each calcium source, we can reproduce a wide range of learning and plasticity protocols, such as Hebbian and anti-Hebbian learning, frequency-dependent plasticity, and unsupervised recognition of frequently repeating input patterns. Moreover, by devising simple neural circuits to provide supervisory signals, we show how the calcitron can implement homeostatic plasticity, perceptron learning, and BTSP-inspired one-shot learning. Our study bridges the gap between theoretical learning algorithms and their biological counterparts, not only replicating established learning paradigms but also introducing novel rules.

ENHANCING STUDENTS’ READING COMPREHENSION OF RECOUNT TEXT WITH MULTIMODAL DIGITAL LITERACY

This study aims firstly to explain the use of multimodal digital literacy to boost students’ reading comprehension of recount text. Secondly, to describe students’ engagement in joining this learning model. It employed a qualitative method. Classroom action research with two cycles was used as the research design. Each cycle consisted of planning, acting, observing, and evaluating. The participants were thirty-ninth graders at the Islamic secondary school in Semarang Regency in the academic year 2024-2025. The pre-test, posttests, and observation guidelines were provided as data collection techniques. The results show that incorporating a genre-based approach and multimodal digital literacy can enhance students’ reading comprehension of recount text. Based on students’ achievement in the pretest, they reached 69, in post-test cycle 1, they achieved 75, and in post-test cycle 2, they obtained 84. Because students were intrigued by multimodal digital literacy, they were also delighted in putting this learning approach into practice. Their grasp of reading comprehension of recall texts was enhanced by the teacher. Together with their peers and the teacher, they studied the course materials. It can be concluded, this learning paradigm improves group work and interpersonal skills, raises learning motivation, and promotes active learners' involvement.Keywords: digital literacy, multimodal; reading comprehension; recount text; revolutionizing.

Privacy-Preserving Incentive Allocation for Fair and Resilient Data Sharing in Resource-Constrained Edge Computing Networks

Efficient and secure data sharing is paramount for advancing modern digital ecosystems, especially within edge computing environments characterized by resource-constrained nodes and dynamic network topologies. In such settings, privacy preservation, computational efficiency, and system resilience are critical for user engagement and overall system performance. However, existing approaches face three primary challenges: (i) limited optimization of privacy protection and absence of dynamic privacy budget scheduling for resource-constrained scenarios, (ii) static incentive mechanisms that overlook individual differences in data quality and resource consumption, and (iii) inadequate strategies to ensure resilience in environments with limited resources and unstable networks. This paper introduces the Federated Learning-based Dynamic Incentive Allocation Framework (FL-DIAF) to address these issues. FL-DIAF integrates differential privacy into the federated learning paradigm deployed on edge nodes, enabling collaborative model training that safeguards individual data privacy while maintaining computational efficiency and system resilience. Additionally, the framework employs a Shapley value-based dynamic incentive allocation model to ensure equitable and transparent distribution of incentives by accurately quantifying each participant’s contribution within an elastic edge computing infrastructure. Comprehensive experimental evaluations on diverse datasets demonstrate that FL-DIAF achieves a 9.573% reduction in the objective function value under typical conditions and attains a 100% task completion rate across all tested resilient edge scenarios.

The Progress and Prospects of Data Capital for Zero-Shot Deep Brain–Computer Interfaces

The vigorous development of deep learning (DL) has been propelled by big data and high-performance computing. For brain–computer interfaces (BCIs) to benefit from DL in a reliable and scalable manner, the scale and quality of data are crucial. Special emphasis is placed on the zero-shot learning (ZSL) paradigm, which is essential for enhancing the flexibility and scalability of BCI systems. ZSL enables models to generalise from limited examples to new, unseen tasks, addressing data scarcity challenges and accelerating the development of robust, adaptable BCIs. Despite a growing number of BCI surveys in recent years, there is a notable gap in clearly presenting public data resources. This paper explores the fundamental data capital necessary for large-scale deep learning BCI (DBCI) models. Our key contributions include (1) a systematic review and comprehensive understanding of the current industrial landscape of DBCI datasets; (2) an in-depth analysis of research gaps and trends in DBCI devices, data and applications, offering insights into the progress and prospects for high-quality data foundation and developing large-scale DBCI models; (3) a focus on the paradigm shift brought by ZSL, which is pivotal for the technical potential and readiness of BCIs in the era of multimodal large AI models.

Targeting the neural bases of novel word acquisition using theta-band transcranial alternating current stimulation

ABSTRACT Novel word acquisition is crucial for both first language acquisition and foreign language learning. Most previous studies of its neural bases were conducted via neuroimaging methods, which can reveal correlational but not causal links between neural and cognitive processes. This paper is one of the first to study the effect of transcranial alternating current stimulation on novel word acquisition. We developed a new explicit word learning paradigm with 80 pseudowords paired with 80 Russian nouns. Online validation revealed no floor or ceiling effects, so the paradigm provides optimal difficulty for potentially modulating performance with tACS. We applied theta-band (5.5 Hz) in-phase tACS over the left frontal-temporal regions in 30 healthy young participants using the paradigm. TACS did not significantly modulate any performance measures and further research is required to test its spatial, frequency, and phase-specific effects.