Dynamic Learning Research Articles

Novelty triggers time-dependent theta oscillatory dynamics in cortical-hippocampal-midbrain circuitry

Rapid adaptation to novel environments is crucial for survival, and this ability is impaired in many neuropsychiatric disorders. Understanding neural adaptation to novelty exposure therefore has therapeutic implications. Here, I found that novelty induces time-dependent theta (4-12Hz) oscillatory dynamics in brain circuits including the medial prefrontal cortex (mPFC), ventral hippocampus (vHPC), and ventral tegmental area (VTA), but not dorsal hippocampus (dHPC), as mice adapt to a novel environment. Local field potential (LFP) recordings were performed while mice were freely behaving in a novel or a familiar arena for 10 min. Initially, mice exhibited increased exploratory behavior upon exposure to novelty, which gradually decreased to levels observed in mice exposed to the familiar arena. Over the same time course, the mPFC, vHPC, and VTA displayed progressively increasing theta power through novelty exposure. Additionally, theta coherence and theta phase synchrony measures demonstrated that novelty weakened the connectivity between these areas, which then gradually strengthened to the level observed in the familiar group. Conversely, mice exposed to the familiar arena showed steady and consistent behavior as well as theta dynamics in all areas. Treatment with a dopamine D1-receptor (D1R) antagonist in the vHPC disrupted neurophysiological adaptation to novelty specifically in the vHPC-mPFC and vHPC-VTA circuits, without affecting behavior. Thus, novelty induces distinct theta dynamics that are not readily dictated by behavior in the mPFC, vHPC, and VTA circuits, a process mediated by D1Rs in the vHPC. The observed time-dependent circuit dynamics in the key learning and memory circuit would provide new insights for treating neuropsychiatric disorders that often show impaired novelty processing.

Phase-amplitude reduction-based imitation learning

In this study, we propose the use of the phase-amplitude reduction method to construct an imitation learning framework. Imitating human movement trajectories is recognized as a promising strategy for generating a range of human-like robot movements. Unlike previous dynamical system-based imitation learning approaches, our proposed method allows the robot not only to imitate a limit cycle trajectory but also to replicate the transient movement from the initial or disturbed state to the limit cycle. Consequently, our method offers a safer imitation learning approach that avoids generating unpredictable motions immediately after disturbances or from a specified initial state. We first validated our proposed method by reconstructing a simple limit-cycle attractor. We then compared the proposed approach with a conventional method on a lemniscate trajectory tracking task with a simulated robot arm. Our findings confirm that our proposed method can more accurately generate transient movements to converge on a target periodic attractor compared to the previous standard approach. Subsequently, we applied our method to a real robot arm to imitate periodic human movements.

A novel dynamic machine learning-based eXplainable fusion monitoring: Application to industrial and chemical processes

Abstract The complexity and fusion dynamism of the modern industrial and chemical sectors have been increasing with the rapid progress of IR 4.0 to 5.0. The transformative characteristics of Industry 4.0 to 5.0 have not been fully explored in terms of the fundamental importance of explainability. Traditional monitoring techniques for automatic anomaly detection, identifying the potential variables, and root cause analysis for fault information are not intelligent enough to tackle the intricate problems of real-time practices in the industrial and chemical sectors. This study presents a novel dynamic Machine Learning (ML) based eXplainable fusion approach to address the issues of process monitoring in industrial and chemical process systems. The methodology aims to detect faults, identify their key causes and feature variables, and analyze the root path of fault propagation with the time and magnitude of one cause variable to another impact. This study proposed using a time domain multivariate granger-entropy-aided dynamic independent component analysis (ICA) - distributed canonical correlation analysis (CCA) approach, incorporating the dynamics time wrapping supported time delay-signed directed graph. The proposed methodology utilized the application to industrial and chemical processes and verified using the Continuous Stirred Tank Reactor (CSTR) and Tennessee Eastman Process (TEP) as practical application benchmarks. The framework's validations and efficiency are evaluated using established techniques such as classic computed ICA and Dynamic ICA as standard model scenarios. The outcomes and results showed that the newly developed strategy is preferable to previous approaches regarding explainability and robust detection and identification of the actual root causes with high FDRs and low FARs.

Adltformer Team-Training with Detr: Enhancing Cattle Detection in Non-Ideal Lighting Conditions Through Adaptive Image Enhancement

This study proposes an image enhancement detection technique based on Adltformer (Adaptive dynamic learning transformer) team-training with Detr (Detection transformer) to improve model accuracy in suboptimal conditions, addressing the challenge of detecting cattle in real pastures under complex lighting conditions—including backlighting, non-uniform lighting, and low light. This often results in the loss of image details and structural information, color distortion, and noise artifacts, thereby compromising the visual quality of captured images and reducing model accuracy. To train the Adltformer enhancement model, the day-to-night image synthesis (DTN-Synthesis) algorithm generates low-light image pairs that are precisely aligned with normal light images and include controlled noise levels. The Adltformer and Detr team-training (AT-Detr) method is employed to preprocess the low-light cattle dataset for image enhancement, ensuring that the enhanced images are more compatible with the requirements of machine vision systems. The experimental results demonstrate that the AT-Detr algorithm achieves superior detection accuracy, with comparable runtime and model complexity, reaching 97.5% accuracy under challenging illumination conditions, outperforming both Detr alone and sequential image enhancement followed by Detr. This approach provides both theoretical justification and practical applicability for detecting cattle under challenging conditions in real-world farming environments.

Dynamic event-triggered data-driven iterative learning consensus control for nonlinear MASs with unknown disturbance

This work tackles the consensus problem for nonlinear nonaffine multi-agent systems (MASs) with unknown disturbances using a data-driven iterative learning control (DDILC) scheme. The unknown model of the agent is first converted through dynamic linearisation into a equivalent data model, comprising a linear combination of an unknown dynamic parameter and an uncertain term. An adaptive algorithm is then formulated to estimate the unknown dynamic parameters accurately. The total uncertainty, which arises from the parameter estimation error and the uncertain term, is estimated by the extended state observer. The estimations are used by the controller to compensate for total uncertainty. Further, considering the resource limitations of communication and the constraints of the actual control input, a constrained iterative learning controller with a dynamic event-triggered mechanism is developed, which guarantees control performance and reduces the number of communications in MASs. It is proven that the tracking error asymptotically approaches a small bound around zero under the proposed DDILC framework. The simulation results confirm the effectiveness of the theoretical research.

The Students’ Perception of Using Visual Literacy in English Language Learning: A Descriptive Study

English as a global lingua franca presents challenges for students in understanding the language in Indonesia, where English is taught as a foreign language. This study aims to explore students' perceptions of the use of visual literacy in English language learning. The use of visual literacy, such as images and videos, has been shown to help improve students' comprehension and engagement in the learning process. The research method used is a descriptive qualitative approach involving seventh-grade students at a middle school in Jember, Indonesia. Data were collected through observations and interviews. The results showed that visual literacy has a positive impact on students' understanding of English language material. Students find it easier to grasp concepts when supported by visual media. Additionally, the use of visual literacy also increases students' motivation and interest in learning, making the learning process more interactive and enjoyable. In the independent curriculum, English language learning encompasses six skills: listening, speaking, reading, writing, reflecting, and presenting, all of which can be supported by the use of visual literacy. Visual literacy not only facilitates the understanding of vocabulary but also encourages students to think critically and creatively through visual interpretation. This creates a more dynamic learning environment and fosters active participation from students. The conclusion of this study is that visual literacy can be an effective tool in enhancing the quality of English language learning, particularly in helping students to understand or remember vocabulary and enhancing students’ English language skills.

Abstract A015: Design and deciphering of precision peptide inhibitors for cancer stemness using generative deep learning and molecular dynamics simulations

Abstract The use of artificial intelligence (AI) and machine learning (ML) for exploring vast amino acid sequence spaces has recently gained traction in the discovery of antibiotics and the design of biomaterials with a wide array of algorithms. However, despite some advancements, designing peptide inhibitors to specifically modulate protein-protein interactions remains a significant challenge. In this contribution, we explore the use of a Long Short-Term Memory (LSTM) network - a type of recurrent neural network - to model peptide sequences, given its ability to process sequential data and capture long-term dependencies, critical for peptide design. Our research focuses on developing precision peptides that target the interaction between the Notch intracellular domain (NICD) and CBF1/RBPJ transcription factors, key regulators of the Notch signaling pathway implicated in breast and pancreatic cancer stemness. We inferred hydrophobic, hydrophilic, van der Waals, and salt bridge interactions from experimentally determined three-dimensional protein complex structures, which were used for feature engineering via one-hot encoding. Peptides, each 20 amino acids in length, were generated using temperature scaling in the LSTM model. These peptides were then structurally optimized and subjected to molecular dynamics (MD) simulations, followed by molecular mechanics Poisson-Boltzmann surface area (MM/PBSA) analysis to assess their interactions and binding affinities with the Notch receptor. The MD simulations provide valuable molecular-level insights into the peptide-Notch interactions, helping to evaluate their binding strength. Further biological testing is underway to validate the efficacy of these lead peptide inhibitors and elucidate their molecular mechanisms in targeting cancer stem cells associated with breast cancer. Citation Format: Gurudeeban Selvaraj, Satyavani Kaliamurthi, Gilles H. Peslherbe. Design and deciphering of precision peptide inhibitors for cancer stemness using generative deep learning and molecular dynamics simulations. [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Optimizing Therapeutic Efficacy and Tolerability through Cancer Chemistry; 2024 Dec 9-11; Toronto, Ontario, Canada. Philadelphia (PA): AACR; Mol Cancer Ther 2024;23(12_Suppl):Abstract nr A015

Application of Video Based Learning Method to Increase Student Engagement in the Digital Age

The rapid development of digital technology has revolutionized the educational landscape, which demands more innovative and engaging teaching methods. One such method is Video-Based Learning (VBL), which utilizes multimedia resources, such as videos, to enhance students' learning experience. This article discusses the application of VBL as a strategy to enhance student engagement in the classroom in the context of the digital age. The research examines the effectiveness of VBL in encouraging active participation, enhancing understanding and strengthening students' motivation to learn. By integrating video content in the learning process, educators can provide a dynamic and interactive learning environment that suits various learning styles. In addition, VBL facilitates access to resources beyond traditional teaching, allowing students to interact with learning materials according to their pace and preferences. This research demonstrates the positive impact of VBL on student engagement, especially in subjects that benefit from visual and auditory stimulation, and discusses challenges and solutions in effective implementation. This article aims to provide insights and practical guidelines for educators in integrating Video-Based Learning in their teaching practices, so that students are better equipped to thrive in the digital age.

Applications and Progress of Machine Learning Techniques in the Ladle Furnace Refining Process: A Review

Machine learning techniques have now become key technologies in optimizing the ladle furnace (LF) refining process, significantly improving prediction accuracy and operational efficiency in various aspects such as temperature control, slag optimization, and alloying processes. This article first reviews the principles of neural networks, utilizing their significant advantages in data processing capability and model generalization to explore the current applications and development trends of machine learning techniques in the LF refining process. It analyzes the advantages of hybrid dynamic models and ensemble learning methods in improving prediction accuracy and handling complex data. Additionally, this article explains the optimization of the LF refining process by combining machine learning techniques with traditional methods to reduce costs and improve production efficiency. It demonstrates the application effects and potential value of machine learning techniques in actual production, discusses the strategies and challenges in real‐time control and prediction accuracy, and envisions the future development direction of intelligent models for the LF refining process, providing a scientific basis and technical reference for the subsequent development of LF refining technology.

Combination of dynamic TOPMODEL and machine learning techniques to improve runoff prediction

AbstractTOPMODEL has been widely employed in hydrology research, undergoing continuous modifications to broaden its practical applicability and enhance its simulation accuracy. To encompass spatial discretization, diffusion‐wave characteristics, depth‐dependent flow velocity, and flux estimation in the unsaturated zone, a generalized dynamic TOPMODEL is developed by introducing a greater number of physical parameters. The present study aims to evaluate the optimal combination of these parameters within the dynamic TOPMODEL framework using machine learning techniques to improve the accuracy of runoff predictions and bolster the model's reliability. An innovative training method is suggested to elevate the model's performance by integrating the Long Short‐Term Memory (LSTM) algorithm and a topological classification, which relies on the evolving spatial distribution of runoff conditions during floods. The research findings show that the proposed methodology achieves the lowest mean relative error (MRE) at 0.106, the highest Pearson correlation coefficient (PC) at 0.938, and the highest coefficient of determination (R2) at 0.906 among the three dynamic TOPMODEL types adopted in this study. The effective implementation of a case study in a river basin showcases the feasibility of the proposed method in conjunction with dynamic TOPMODEL and underscores the importance of employing the suggested training procedure.

Enhancing low-illumination imagery using a Deep Convolutional Generative Adversarial Network with weight regularization (DCGAN-WR) and Zero-Reference Deep Curve Estimation (DCE)

Improving low-light images to enhance prediction in various applications has a greater advantage. A two-pronged approach that employs Deep Convolutional Generative Adversarial Networks with weight regularization (DCGAN-WR) and Zero-Reference Deep Curve Estimation (DCE) was used. The model was trained using the LOL dataset, and the results showed significant improvements in image quality. The DCGAN is fine-tuned with Group Lasso regularization to enhance the performance. The DCGAN-WR model is shown to enhance images realistically, demonstrating its capacity to learn characteristics and texture representations from low-light input. Empirical and simulated image comparisons demonstrate remarkable performance under demanding low-light settings. Moreover, the DCE model employs a novel approach that considers color constancy loss, light smoothness, and spatial consistency. Information about the learning dynamics and curve parameter changing capabilities of the model can be visualized by loss function graphs, which aim to maximize picture quality. Compared to the original images, Images generated by the DCE models maintain color accuracy, increase exposure levels, and preserve spatial coherence. A solution for low-illumination image enhancement is achieved through the proposed model DCGAN-WR and DCE. Genuine details are recorded by the GAN model, while the DCE adjusts the exposure levels and color balance to produce improved, aesthetically pleasing, and contextually accurate images. The proposed approach not only outperforms the other methods on the LOL dataset but also exhibits potential for practical use in computer vision tasks, which require higher image quality for precise analysis and interpretation.

Exploring the Scalability of 5G-Enabled E-Learning Platforms in Rural Areas

Background. The rise of 5G technology offers the potential to revolutionize e-learning, particularly in rural areas where access to high-quality educational resources has historically been limited. Despite the promise of improved connectivity, the scalability and impact of 5G-enabled e-learning platforms in these regions remain underexplored. Purpose. This study aims to assess the feasibility and scalability of implementing 5G-enabled e-learning platforms in rural areas and to evaluate their effectiveness in enhancing educational access and outcomes. Method. The research employs a mixed-method approach, combining quantitative analysis of internet connectivity, user engagement, and learning outcomes with qualitative interviews from educators and students. Data was collected from five rural regions across three countries where 5G infrastructure was recently introduced. The study measured improvements in e-learning adoption rates, student engagement, and academic performance over a 12-month period. Results. Results indicate that 5G-enabled e-learning platforms significantly improved both access to education and student engagement, with a 30% increase in platform adoption and a 25% improvement in student performance. Educators also reported more dynamic and interactive learning experiences. However, challenges related to infrastructure costs and digital literacy were noted, which could limit scalability in certain regions. Conclusion. In conclusion, the study demonstrates the potential of 5G technology to scale e-learning platforms in rural areas, improving educational access and outcomes. Nonetheless, addressing challenges such as infrastructure costs and digital literacy is critical for ensuring long-term scalability and success.

Personalized College English Learning Experience Assisted by Artificial Intelligence: An Algorithm-Driven Adaptive Learning Approach

Personalized College English Learning offers customized education based on individual necessities and aims. Students enhance their linguistic ability efficiently over engaging sessions, specific curricula, and professional teachers. This study aims to progress a smart Artificial Intelligence (AI)-based adaptive learning method for enhancing personalized college English learning experience. Our proposed model provides an intelligent detecting device-enabled setting for learning English over big data analysis and machine learning based on a DM approach. Intelligent sensing strategies detention significant statistics from college English students and big data analysis examines the data to produce beneficial insights that increase real-time personalizing of English learning experiences. We suggest an innovative Starling Murmuration fine-tuned Dynamic Weighted Random Forest (SM-DWRF) system for classifying the data. Our model influences perceptions from collective behavior in starling flocks to dynamically alter feature weights in DWRF arrangement. It integrates fine-tuning mechanisms to adaptively weight features, developing classification accurateness and strength in adapted college English learning settings. We implemented our suggested model in Python software. In the assessment stage, we accurately measure the effectiveness of our proposed SM-DWRF model in identifying diverse features of English learning across various parameters. Our experimental findings incontestably showcase the greater performance of our model associated with conventional approaches in classifying content from multimodal statistics. Significantly, we perceive prominent enhancements in reliability and robustness, when acclimating to dynamic learning settings.

The Role of Mobile Phones in Facilitating Business Education During the Transition to New Normal Learning Models

This study explores the impact of mobile phone use on self-regulated learning among students at the Nueva Ecija University of Science and Technology San Isidro Campus, specifically within the College of Management and Business Technology. Using a descriptive correlational research design, the study aims to assess how integral mobile phones are to students’ learning processes, focusing on activities such as goal setting, environment structuring, task strategies, time management, help-seeking, and self-evaluation. Data were collected through a survey distributed to a significant portion of the student body, with subsequent statistical analysis to elucidate the relationship between mobile phone use and the efficacy of self-regulated learning strategies. Results reveal that mobile phones play a crucial role in enhancing students’ ability to manage their learning autonomously. They facilitate access to resources, enable communication with peers and instructors, and support various aspects of self-regulated learning. However, the study also identifies challenges such as distraction, dependency, and varying access to digital resources, suggesting a need for structured guidance in mobile usage. The assumptions emphasize the potential of mobile phones as beneficial educational tools when integrated with supportive measures to mitigate associated challenges. This study contributes to the understanding of mobile learning dynamics and underscores the need for educational strategies that leverage mobile technology effectively.

Dynamic temporal reinforcement learning and policy-enhanced LSTM for hotel booking cancellation prediction

Dynamic temporal reinforcement learning and policy-enhanced LSTM for hotel booking cancellation prediction

The Hard Pursuit of Optimal Vaccination Compliance in Heterogeneous Populations

Abstract The present model focuses on how people decide to get vaccinated, based on their beliefs and costs and on the public information concerning the disease severity, that in turn depends on the vaccination coverage. This interplay between beliefs and outcomes gives rise to a dynamical learning process, whose steady state is a self-fulfilling equilibrium. Although equilibrium levels of immunization and disease severity do not depend on beliefs, even in equilibrium heterogeneous people interpret the severity/coverage relation in different ways. These differences, together with the structural parameters of the model, have important implications for the stability of the equilibrium, finally impinging on the efficacy of policies aimed at correcting the existing state. In fact, we find that when the equilibrium disease severity is too high and immunization is suboptimal, mandatory vaccination and nudging can be valid options for fighting vaccination hesitancy (while moral suasion looks less effective); in addition, some policy mixes turn out to be very effective. However, given the interaction between beliefs, learning, and policies, the wished policy goal cannot be achieved immediately with precision, due to possible instability of equilibria. This supports the idea that immunization must be encouraged by using mixes of persistent policies.

A robotic skill transfer learning framework of dynamic manipulation for fabric placement

Placing fabric poses a challenge to robots since fabric with high dimensional configuration space can deform during manipulation. Existing methods for placing fabric mostly rely on static operations, which are inefficient and require a large workspace. Therefore, this study applies dynamic manipulation (manipulating uncontrollable parts of the fabric by swinging) to fabric placement, proposing a novel learning framework for robotic dynamic fabric placement skill learning and generalization. The proposed framework integrates reinforcement learning with imitation learning, leveraging expert demonstration data to guide and accelerate skill acquisition. Additionally, fabric characteristics are combined with imitation learning to enable the transfer and generalization of the learned policy to real-world environments The experiments suggest that the proposed framework is capable of achieving the placement tasks for a range of positions and fabrics. For success rate, the policy of the proposed framework ultimately achieves a flatness of exceeding 95% and a placement distance error of less than 2 mm. Moreover, the proposed approach is similar in operation time to the fastest method, while it can reduce the space required for manipulating the fabric by over 15%. Compared with other placement policies, it is promising because of its high accuracy, flexibility, efficiency, as well as adaptability.

Social Dynamics and Language Learning: Do Extroverts Excel in L2 Acquisition? An Analysis through the Lens of Interactionism Theory

The objective of this study is to explore the relationship between social dynamics and second language acquisition, with a specific emphasis on whether extroverts surpass in learning English as a second language more efficiently and effectively than introverts. Moreover, the research is grounded in the Interactionism theory by Lev Vygotsky (1962) which states that language development is hugely determined by social and cultural communication and interaction. So, this study examines whether the personality traits such as being extrovert affect learner’s engagement to learn language more efficiently by socially interacting with other people or not. For this purpose, students of the BS English department of Army Burn Hall College for Girls, Abbottabad were selected as the participants to share their views through a mixed-method approach. A structured questionnaire was designed to gather insights from the students. Results for the close-ended questions have been shown through graphs whereas thematic analysis has been done for open-ended questions. The findings revealed that though ESL learners face many challenges during social meet-ups they still learn vocabulary, pronunciation, confidence, fluency, etc.

Tucker Decomposition Enhanced Dynamic Graph Convolutional Networks for Crowd Flows Prediction

Crowd flows prediction is an important problem for traffic management and public safety. Graph Convolutional Network (GCN), known for its ability to effectively capture and utilize topological information, has demonstrated significant advancements in addressing this problem. However, GCN-based models were often based on predefined crowd-flow graphs via historical movement behaviors of human beings and traffic vehicles, which ignored the abnormal changes in crowd flows. In this study, we propose a multi-scale fusion GCN-based framework with Tucker decomposition named mTDNet to enhance dynamic GCN for Crowd flows prediction. Following the paradigm of extant methods, we also employ the predefined crowd-flow graphs as a part of mTDNet to effectively capture the historical movement behaviors of crowd flows. To capture the abnormal changes, we propose a Tucker decomposition-based network with the product of the adjacency matrix of historical movement pattern graphs and an adaptive learning tensor ( \(ALT\) ) by reconstructing the crowd flows. Particularly, we utilize the Tucker decomposition scheme to decompose \(ALT\) , which enhances the dynamic learning of graph structures, allowing for effective capturing of the dynamic changes in crowd flow, including abnormal changes. Furthermore, a multi-scale three-dimensional GCN is utilized to mine and fuse the multiscale spatio-temporal information from crowd flows, to further boost the mTDNet prediction performance. Experiments conducted on two real-world datasets showed that the proposed mTDNet surpasses other crowd flow prediction methods.

TMODINET: A trustworthy multi-omics dynamic learning integration network for cancer diagnostic

Multiple types of omics data contain a wealth of biomedical information which reflect different aspects of clinical samples. Multi-omics integrated analysis is more likely to lead to more accurate clinical decisions. Existing cancer diagnostic methods based on multi-omics data integration mainly focus on the classification accuracy of the model, while neglecting the interpretability of the internal mechanism and the reliability of the results, which are crucial in specific domains such as precision medicine and the life sciences. To overcome this limitation, we propose a trustworthy multi-omics dynamic learning framework (TMODINET) for cancer diagnostic. The framework employs multi-omics adaptive dynamic learning to process each sample to provide patient-centered personality diagnosis by using self-attentional learning of features and modalities. To characterize the correlation between samples well, we introduce a graph dynamic learning method which can adaptively adjust the graph structure according to the specific classification results for specific graph convolutional networks (GCN) learning. Moreover, we utilize an uncertainty mechanism by employing Dirichlet distribution and Dempster-Shafer theory to obtain uncertainty and integrate multi-omics data at the decision level, ensuring trustworthy for cancer diagnosis. Extensive experiments on four real-world multimodal medical datasets are conducted. Compared to state-of-the-art methods, the superior performance and trustworthiness of our proposed algorithm are clearly validated. Our model has great potential for clinical diagnosis.