- Research Article
- 10.1007/s10950-026-10387-3
Proposing an optimal input feature combination for estimating ground response spectra using machine learning: a study in New Zealand
- Mar 28, 2026
- Journal of Seismology
- Le-Anh-Nhat Nguyen + 3 more +3
Publications from 2021 to 2026
Showing 10 of 292 papers
Proposing an optimal input feature combination for estimating ground response spectra using machine learning: a study in New Zealand
AI-Driven inverse design of 3D-printed continuous fiber composites using multiscale homogenization and dual neural networks
Assessing machine learning strategy approaches for early warning of liquidity risk
Artificial intelligence (AI) and machine learning (ML) have been increasingly adopted in the banking sector due to their ability to analyze large-scale datasets, process complex variables, and uncover hidden patterns, especially in the context of liquidity risk, posing a significant challenge for commercial banks. This study contributes to the field by conducting a comprehensive evaluation of several widely used early warning models, such as least absolute shrinkage and selection operator (LASSO) regression, random forest (RF), and extreme gradient boosting (XGBoost), to identify the most suitable approach for forecasting liquidity risk in Vietnamese commercial banks (VCBs) based on VCBs data over the period of 2014–2023. By pinpointing key indicators associated with liquidity crises, these models can assist banks and regulatory authorities in implementing timely preventive measures and enhancing risk management strategies. As a result, the RF model outperforms other methods in identifying possible liquidity crises, according to the empirical results, with an accuracy rate of 99.8 percent. These findings provide bank managers and policymakers with a powerful tool for timely preventive measures, thereby enhancing the resilience and stability of the financial system.
Read moreCost-efficient seismic retrofit strategy for AFRP-retrofitted RC piloti frames using ML-driven plastic hinge modeling
The Art of Incremental Dominance
This chapter examines China's comprehensive strategy in the SCS through the theoretical lens of realism in international relations. Employing qualitative document analysis, the research identifies four integrated strategic pillars: incremental “salami slicing” tactics that gradually alter the status quo without triggering military confrontation; unprecedented artificial island construction and militarization establishing strategic outposts; systematic deployment of coast guard and maritime militia forces operating in legal gray zones; and legal-diplomatic strategies linked with the Belt and Road Initiative. The findings reveal that China's actions exemplify revisionist power behavior in an anarchic system, utilizing coordinated instruments to maximize national interests while avoiding direct conflict. This analysis provides crucial insights for policymakers regarding the multidimensional nature of great power competition and the necessity for comprehensive, long-term regional response strategies.
Read moreExploring the landscape of digital learning materials in education: A PRISMA-based review
In recent years, digital learning materials have increasingly replaced and complemented traditional resources, and they are gradually being standardized to ensure higher levels of accuracy and diversity. Digital educational resources have become an indispensable component of modern education, offering numerous advantages over traditional materials. They allow learners to access content anytime and anywhere, and they can vividly illustrate mechanisms, processes, and natural phenomena that conventional resources often fail to convey effectively. However, the growing integration of digital materials in education also presents significant challenges that require further consideration. Despite growing research interest, comprehensive overviews remain limited. This study aims to address this gap by systematically analysing recent research trends, influential publications, thematic focuses, and future directions in digital learning materials. A systematic review was conducted via the PRISMA framework, with Scopus selected as the primary database for rigorous quality control. The inclusion criteria included peer-reviewed journal articles in English published from 2017 onwards containing “digital learning material” or “digital teaching material” in the title, abstract, or keywords. From an initial pool of 404 articles, 45 met the final selection criteria. The data were cleaned, coded, and analysed via VOSviewer for keyword co-occurrence and network mapping. The findings reveal significant growth in publications, especially after 2020, driven by the global shift toward online learning. Sixteen thematic clusters were identified, with collaborative learning and mathematics serving as key connectors. Fifteen categories of challenges emerged, including limited generalizability and small sample sizes. Five major research directions were proposed: spanning methodology, contextualization, technology integration, teacher training, and assessment. This study provides the first systematic, PRISMA-based synthesis of digital learning materials research, offering a holistic framework to guide future scholarly and practical advancements in the field.
Read moreFault-tolerant trajectory tracking of mobile robots via model-free intelligent PID and fault observer
This paper addresses the problem of unexpected errors, such as model uncertainties and actuator faults, that degrade the performance of wheeled mobile robots (WMR). To overcome these challenges, a fault-tolerant control (FTC) approach is developed in which model-free control (MFC) is merged with an intelligent-proportional integral derivative (i-PID) controller and complemented by a fault observer (FO). Unlike existing approaches, the proposed controller does not rely on accurate system modeling; instead, MFC ensures robustness to time-varying parameters, and while i-PID enhances trajectory tracking through adaptive gain adjustment. The FO estimates actuator faults in real time and compensates for their effects, ensuring reliable operation even under severe conditions. The closed-loop stability is rigorously analyzed via Lyapunov theory. MATLAB/Simulink results show reduced tracking errors, improved stability, and strong robustness under both model uncertainties and actuator faults, including time-varying mass and inertia, validating the effectiveness and practical potential of the proposed FTC.
Read moreGlobal Sliding Mode Fault-Tolerant Control for Nonholonomic Wheeled Robots: Ensuring Prescribed Performance in the Presence of Faults and Slippage
This paper proposes a global integral sliding mode control (GISMC) scheme integrated with an improved disturbance observer (IDO) for robust trajectory tracking of nonholonomic wheeled mobile robots under uncertainties, disturbances, wheel slippage, and actuator faults. The NWMR dynamics are first formulated, followed by the development of the IDO to estimate and compensate for lumped disturbances in real time. The GISMC framework is then designed to ensure finite-time convergence and robustness, while a prescribed performance function (PPF) is incorporated to constrain the tracking errors within predefined bounds and guarantee system stability under dynamic conditions. Experiment results confirm that the proposed GISMC maintains tracking errors within ±0.012 m significantly smaller than those of the backstepping–adaptive sliding mode control (≈0.043 m) and achieves faster convergence with smoother control action than the backstepping–SMC with IDO. The results validate that the proposed controller adapts effectively in real time, providing enhanced robustness and precision in complex and uncertain environments. These findings highlight the potential of the proposed GISMC for practical applications in autonomous mobile systems requiring high accuracy and fault tolerance.
Read moreImpact of surfactant concentration on applicability of MnFe2O4@poly(acrylic) acid nanoparticles in MRI and hyperthermia
A Spectral Clustering-Based Approach for Balancing Data in TF-Target Gene Interaction Prediction Using Heterogeneous Network Embedding.
Identifying transcription factors (TFs) interacting with target genes plays a crucial role in regulating gene expression, providing deep insights into the molecular mechanisms involved in biological processes and diseases, while also opening new opportunities for research and treatment, particularly in precision medicine. However, challenges arise from the high cost of experiments and biological complexity. Recent machine learning models show promise for predicting TF target gene interactions but struggle with data imbalance, as the number of known interactions is much smaller than unknown ones, reducing prediction performance. To address this challenge, we propose an approach that integrates sample selection from potential unknown interactions with spectral clustering to balance the dataset. We construct an adjacency matrix from known interactions, apply spectral clustering to partition the data, and select unknown interactions farthest from the cluster center for diversity. A deep learning method combining random walk sampling and skip-gram embeddings is used to learn representations of biological networks. Experimental results, evaluated via five-fold cross-validation, show an average AUC of 0.9575 ± 0.0044, outperforming existing methods. These findings validate our approach's efficacy in improving prediction accuracy and resolving data imbalance, offering a robust framework for discovering novel TF-target gene interactions in molecular biology and biomedical research.
Read more