Search Results (306)

Bayesian optimization (BO) struggles with data scarcity and poor scalability in high-dimensional many-objective optimization problems. To address this, we propose Co-Evolutionary GAN–Bayesian Optimization (CoEGAN-BO), a novel framework that synergizes generative adversarial networks (GANs) with Bayesian co-evolutionary search for data-driven optimization. The GAN module generates synthetic samples conditioned on promising regions identified by BO, while a co-evolutionary mechanism maintains two interacting populations: one explores the GAN’s latent space for diversity, and the other exploits BO’s probabilistic model for convergence. A bi-stage infilling strategy further enhances efficiency: early iterations prioritize exploration via $L_p$-norm-based candidate selection, later switching to a max–min distance criterion for Pareto refinement. Experiments on expensive multi/many-objective benchmarks show that CoEGAN-BO outperforms four state-of-the-art surrogate-assisted algorithms, achieving superior convergence and diversity under limited evaluation budgets. Full article

Rural resilience building has gained increasing scholarly attention, yet existing literature overlooks the temporal dynamics of resilience evolution and lacks an integrative framework to explain cross-level mechanisms. This paper uses a longitudinal case study to explore how rural resilience transitions from a low-equilibrium to a high-equilibrium state and how neo-endogenous practices emerge in a weak institutional context. The study reveals three key findings. First, the village’s resilience evolved through three phases—institutional intervention, community capital activation, and resilience self-reinforcement—driven by co-evolutionary interactions between an enabling government and the rural community. This process is marked by chain effects of multidimensional community capital (e.g., cultural capital enhancing social capital) and overflow effects from resilience amplification (e.g., multi-scalar network). Second, exogenous resources and endogenous community capital are critical in the neo-endogenous model, but their synergy relies on vertical institutional interventions that foster horizontal networks and enhance communities’ resource absorption capacity. Third, the government enables resilience building by creating a support ecosystem that transitions from institutionally bundled resources to a higher-order composite space, facilitated by urban–rural interactions and community restructuring. The study makes three theoretical contributions: (1) it proposes an analytical framework integrating an enabling government, community capital, and ecosystem upgrading, thus advancing beyond the current community capital-centric paradigm; (2) it introduces a three-phase process model that unpacks spatiotemporal interactions across urban-rural interfaces, multi-scalar networks, and state-community relations, addressing the limitations of static factor-based analyses; (3) it reconceptualizes the role of government as an “enabling government” that mediates local and extra-local resource interfaces, challenging the neo-endogenous theories’ neglect of institutional agency. These insights contribute to rural resilience scholarship through a complex adaptive systems lens and offer policy implications for synergistic urban-rural revitalization. Full article

The rapid expansion of the Internet of Things (IoT) has introduced new vulnerabilities that traditional security mechanisms often fail to address effectively. Signature-based intrusion detection systems cannot adapt to zero-day attacks, while rule-based solutions lack scalability for the diverse and high-volume traffic in IoT environments. To strengthen the security framework for IoT, this paper proposes a deep learning-based anomaly detection approach that integrates Convolutional Neural Networks (CNNs) and Bidirectional Gated Recurrent Units (BiGRUs). The model is further optimized using the Moth–Flame Optimization (MFO) algorithm for automated hyperparameter tuning. To mitigate class imbalance in benchmark datasets, we employ Generative Adversarial Networks (GANs) for synthetic sample generation alongside Z-score normalization. The proposed CNN–BiGRU + MFO framework is evaluated on two widely used datasets, UNSW-NB15 and UCI SECOM. Experimental results demonstrate superior performance compared to several baseline deep learning models, achieving improvements across accuracy, precision, recall, F1-score, and ROC–AUC. These findings highlight the potential of combining hybrid deep learning architectures with evolutionary optimization for effective and generalizable intrusion detection in IoT systems. Full article

Against the backdrop of inclusive development and modernization of employment governance, the limitations of traditional approaches to promoting employment for persons with disabilities—such as information asymmetries and inefficient resource allocation—have become increasingly salient. Building a systematic promotion framework for disability employment has therefore emerged as a critical agenda for advancing modern social governance. Drawing on bounded rationality and information asymmetry theories, this study develops a tripartite evolutionary game model encompassing government, employers, and persons with disabilities. By incorporating key elements such as initial intentions, skill matching, and policy signal transmission, the model analyzes the strategic choices and dynamic interactions among stakeholders. We conduct numerical simulations using delay differential equations (DDEs), perform stability and sensitivity analyses in MATLAB R2024b, and triangulate findings with a practice-based case from Shanghai. The results indicate that persons with disabilities exhibit the highest policy responsiveness within the employment ecosystem and act as the core driver of convergence toward desirable equilibria through four mechanisms: skill-matching effects, policy signal diffusion, perceived institutional fairness, and system-level synergy gains. Although employer subsidies and penalties directly target firms, they exert the strongest psychological incentive effects on persons with disabilities, revealing a “misaligned incentives” feature in policy signaling. Systemic synergy gains activate market network effects, facilitating a pivotal shift from “policy transfusion” to “market self-sustenance.” Based on these findings, we propose a diversified policy toolkit, enhanced policy signaling mechanisms, and innovations in concentrated employment models to support the modernization of disability employment governance. Full article

Knowledge Tracing aims to assess students’ mastery of knowledge concepts in real time, playing a crucial role in providing personalized learning services in intelligent tutoring systems. In recent years, researchers have attempted to introduce Neural Architecture Search (NAS) into knowledge tracing tasks to automatically design more efficient network structures. However, existing NAS-based methods for Knowledge Tracing suffer from excessively large search spaces and slow search efficiency, which significantly constrain their practical applications. To address these limitations, this paper proposes an Adaptive Neural Architecture Search framework based on Transformers for KT, called ANT-KT. Specifically, we design an enhanced encoder that combines convolution operations with state vectors to capture both local and global dependencies in students’ learning sequences. Moreover, an optimized decoder with a linear attention mechanism is introduced to improve the efficiency of modeling long-term student knowledge state evolution. We further propose an evolutionary NAS algorithm that incorporates a model optimization efficiency objective and a dynamic search space reduction strategy, enabling the discovery of high-performing yet computationally efficient architectures. Experimental results on two large-scale real-world datasets, EdNet and RAIEd2020, demonstrate that ANT-KT significantly reduces time costs across all stages of NAS while achieving performance improvements on multiple evaluation metrics, validating the efficiency and practicality of the proposed method. Full article

Leishmania major is a unicellular protozoan that causes cutaneous leishmaniasis in mammals and is mainly transmitted by the sand fly Phlebotomus papatasi. However, the contribution of microRNAs (miRNAs) and protein-coding genes to its pathogenic mechanisms remains largely unexplored. In this study, we systematically analyzed miRNAs and protein-coding genes in L. major, its insect vector, and mammalian hosts. Comparative genomic analysis revealed 2963 conserved proteins shared among the three groups, highlighting a core set of proteins across protozoa, vectors, and hosts. Among mammals, human proteins exhibited the highest homology with L. major, while P. papatasi displayed the lowest proportion of homologs. Functional annotation of 94 hypothetical proteins identified 27 infection-related proteins, including 24 protein kinases and three tyrosine phosphatases, which may represent novel therapeutic targets. In addition, an EST-based approach identified 29 novel miRNAs in L. major. Phylogenetic analysis indicated that these miRNAs diverged into two distinct evolutionary branches, and homology analysis revealed that seven miRNAs were absent in all mammalian species. For example, miR-10117-3p was detected only in nematode Heligosmoides polygyrus. Furthermore, miRNA-gene interaction network analysis highlighted four key genes potentially involved in L. major infection. Collectively, our findings expand current knowledge of protozoan virulence by identifying novel miRNAs and infection-related proteins and provide promising candidates for future drug development against leishmaniasis. Full article

Adenylate kinase (ADK), a highly conserved and ubiquitously expressed enzyme in plants, serves as a critical regulator of cellular energy homeostasis and abiotic stress adaptation. While ADK families have been characterized in model species (e.g., Arabidopsis thaliana, Oryza sativa) and crops such as tomato (Solanum lycopersicum), the molecular features and stress-responsive roles of ADK genes in pepper (Capsicum annuum L.) remain uncharacterized. Here, we systematically identified 15 ADK genes in pepper (named by chromosomal location) and revealed their evolutionary relationships with orthologs from four plant species, clustering into six conserved groups. The promoters of CaADKs were found to contain cis-acting elements linked to stress responses, including those responsive to abscisic acid, gibberellin, and low-temperature conditions. Tissue-specific expression profiling highlighted CaADK9 as a ubiquitously expressed member, suggesting a housekeeping function in basal biological processes. Notably, functional assays under low-temperature and salt stress revealed distinct regulatory patterns: CaADK11 and CaADK12 were significantly downregulated, while CaADK9 was upregulated under salt stress, indicating specialized roles in stress signaling. Additionally, we identified ADK-interacting partners involved in nucleotide homeostasis, providing novel insights into the molecular network underlying pepper’s stress responses. This study represents the first comprehensive analysis of the CaADK family, laying a foundation for unraveling ADK-mediated stress adaptation mechanisms in Solanaceous crops. Full article

The precise prediction of the remaining useful life (RUL) of lithium-ion batteries is of great significance for improving energy management efficiency and extending battery lifespan, and it is widely applied in the fields of new energy and electric vehicles. However, accurate RUL prediction still faces significant challenges. Although various methods based on deep learning have been proposed, the performance of their neural networks is strongly correlated with the hyperparameters. To overcome this limitation, this study proposes an innovative approach that combines the Alpha evolutionary (AE) algorithm with a deep learning model. Specifically, this hybrid deep learning architecture consists of convolutional neural network (CNN), time convolutional network (TCN), bidirectional long short-term memory (BiLSTM) and multi-scale attention mechanism, which extracts the spatial features, long-term temporal dependencies, and key degradation information of battery data, respectively. To optimize the model performance, the AE algorithm is introduced to automatically optimize the hyperparameters of the hybrid model, including the number and size of convolutional kernels in CNN, the dilation rate in TCN, the number of units in BiLSTM, and the parameters of the fusion layer in the attention mechanism. Experimental results demonstrate that our method significantly enhances prediction accuracy and model robustness compared to conventional deep learning techniques. This approach not only improves the accuracy and robustness of battery RUL prediction but also provides new ideas for solving the parameter tuning problem of neural networks. Full article

The spread of epidemic-associated panic information through online social platforms, as well as the allocation and utilization of therapeutic defensive resources in reality, directly influences the transmission of infectious diseases. Moreover, how to reasonably allocate resources to effectively suppress epidemic spread remains a problem that requires further investigation. To address this, we construct a coupled three-layer network framework to explore the complex co-evolutionary mechanisms among false panic information, therapeutic defensive resource transmission, and disease propagation. In the model, individuals can obtain therapeutic defensive resources either through centralized distribution by government agencies or through interpersonal assistance, while the presence of false panic information reduces the willingness of neighbors to share resources. Using the microscopic Markov chain approach, we formulate the dynamical equations of the system and analyze the epidemic threshold. Furthermore, systematic simulation analyses are carried out to evaluate how panic information, resource-sharing willingness, centralized distribution strategies, and resource effectiveness affect epidemic prevalence and threshold levels. For example, under a representative parameter setting, the infection prevalence decreases from 0.18 under the random allocation strategy to 0.03 when resources are allocated exclusively to infected individuals. Moreover, increasing the total supply of resources under high treatment efficiency raises the epidemic threshold by approximately 2.5 times, effectively delaying the outbreak. These quantitative results highlight the significant role of allocation strategies, resource supply, and treatment efficiency in suppressing epidemic transmission. Full article

Boll abscission in cotton (Gossypium spp.) is a key factor that limits yield; however, the molecular mechanisms underlying this process remain poorly understood. In this study, boll abscission characteristics were uncovered in four cotton varieties that exhibited extreme differences in boll abscission rates via tissue sectioning. Transcriptome analysis was performed on the four cotton varieties. Using weighted gene co-expression network analysis (WGCNA) of the transcriptome data, we identified a stearoyl-(acyl-carrier-protein) desaturase (SACPD) as a potential key regulator of boll abscission. We also performed evolutionary analyses on the SACPD gene family across five cotton species and identified 63 members that were classified into four evolutionary clades, with duplication-polyploidization events being a major driver of gene expansion. Tissue-specific expression profiling revealed that Gossypium hirsutum GhSACPD19 is highly expressed in the abscission zone. Our findings suggest a role of GhSACPD19 in regulating boll abscission, likely through metabolism of jasmonate, a well-known positive regulator of abscission. Our work offers new insights into the regulation of organ abscission at cellular and molecular levels and presents a valuable resource for cotton yield improvement. Full article

Accurate reactive power optimization in wind farms (WFs) is critical for optimizing operations and ensuring grid stability, yet it faces challenges from noisy, nonlinear, and dynamic Supervisory Control and Data Acquisition (SCADA) data. This study proposes an innovative framework, WBS-BiGRU, integrating three novel components to address these issues. Firstly, the Wavelet-DBSCAN (WDBSCAN) method combines wavelet transform’s time–frequency analysis with density-based spatial clustering of applications with noise (DBSCAN)’s density-based clustering to effectively remove noise and outliers from complex WF datasets, leveraging multi-scale features for enhanced adaptability to non-stationary signals. Subsequently, a Boomerang Evolutionary Optimization (BAEO) with the Seasonal Decomposition Improved Process (SDIP) synergistically decomposes time series into trend, seasonal, and residual components, generating diverse candidate solutions to optimize data inputs. Finally, a Bidirectional Gated Recurrent Unit (BiGRU) network enhanced with an attention mechanism captures long-term dependencies in temporal data and dynamically focuses on key features, improving reactive power forecasting precision. The WBS-BiGRU framework significantly enhances forecasting accuracy and robustness, offering a reliable solution for WF operation optimization and equipment health management. Full article

SHORT INTERNODE (SHI)-Related Sequence (SRS) transcription factors play crucial roles in plant growth, development, and stress responses and have been extensively studied in various plant species. However, the molecular functions and regulatory mechanisms of SRS genes in the economically important tropical fruit crop pitaya (Hylocereus undatus) remain poorly understood. This study identified 9 HuSRS genes in pitaya via bioinformatics analysis, with subcellular localization predicting nuclear distributions for all. Gene structure analysis showed 1–4 exons, and conserved motifs (RING-type zinc finger and IXGH domains) were shared across subclasses. Phylogenetic analysis classified the HuSRS genes into three subfamilies. Subfamily I (HuSRS1–HuSRS4) is closely related to poplar and tomato homologs and subfamily III (HuSRS6–HuSRS8) contains a recently duplicated paralogous pair (HuSRS7/HuSRS8) and shows affinity to rice SRS genes. Protein structure prediction revealed dominance of random coils, α-helices, and extended strands, with spatial similarity correlating to subfamily classification. Interaction networks showed HuSRS1, HuSRS2, HuSRS7 and HuSRS8 interact with functional proteins in transcription and hormone signaling. Promoter analysis identified abundant light/hormone/stress-responsive elements, with HuSRS5 harboring the most motifs. Transcriptome and qPCR analyses revealed spatiotemporal expression patterns: HuSRS4, HuSRS5, and HuSRS7 exhibited significantly higher expression levels in callus (WG), which may be associated with dedifferentiation capacity. In seedlings, HuSRS9 exhibited extremely high transcriptional accumulation in stem segments, while HuSRS1, HuSRS5, HuSRS7 and HuSRS8 were highly active in cotyledons. This study systematically analyzed the characteristics of the SRS gene family in pitaya, revealing its evolutionary conservation and spatio-temporal expression differences. The research results have laid a foundation for in-depth exploration of the function of the SRS gene in the tissue culture and molecular breeding of pitaya. Full article

Dongxiang wild rice (DXWR, Oryza rufipogon Griff.), the northernmost known wild rice species, exhibits exceptional tolerance to combined low-temperature and anaerobic stress during seed germination, providing a unique model for understanding plant adaptation to complex environmental constraints. Here, we employed an integrated multi-omics approach combining genomic, transcriptomic, and metabolomic analyses to unravel the synergistic regulatory mechanisms underlying this tolerance. Genomic comparative analysis categorized DXWR genes into three evolutionary groups: 18,480 core genes, 15,880 accessory genes, and 6822 unique genes. Transcriptomic profiling identified 10,593 differentially expressed genes (DEGs) relative to the control, with combined stress triggering the most profound changes, specifically inducing the upregulation of 5573 genes and downregulation of 5809 genes. Functional characterization revealed that core genes, including DREB transcription factors, coordinate energy metabolism and antioxidant pathways; accessory genes, such as glycoside hydrolase GH18 family members, optimize energy supply via adaptive evolution; and unique genes, including specific UDP-glycosyltransferases (UDPGTs), confer specialized stress resilience. Widely targeted metabolomics identified 889 differentially accumulated metabolites (DAMs), highlighting significant accumulations of oligosaccharides (e.g., raffinose) to support glycolytic energy production and a marked increase in flavonoids (153 compounds identified, e.g., procyanidins) enhancing antioxidant defense. Hormonal signals, including jasmonic acid and auxin, were reconfigured to balance growth and defense responses. We propose a multi-level regulatory network based on a “core-unique-adaptive” genetic framework, centered on ERF family transcriptional hubs and coordinated through a metabolic adaptation strategy of “energy optimization, redox homeostasis, and growth inhibition relief”. These findings offer innovative strategies for improving rice stress tolerance, particularly for enhancing germination of direct-seeded rice under early spring low-temperature and anaerobic conditions, by utilizing key genes such as GH18s and UDPGTs, thereby providing crucial theoretical and technological support for addressing food security challenges under climate change. Full article

The submandibular glands (SMGs), along with the parotid and sublingual glands, generate the majority of saliva and play critical roles in maintaining oral and systemic health. Despite their physiological importance, long-term therapeutic options for salivary gland dysfunction remain limited, highlighting the need for a deeper molecular understanding of SMG biology, particularly in humans. To address this knowledge gap, we have performed transcriptomic- and epigenomic-based analyses and molecular characterization of the human SMG. Our integrated analysis of multiorgan RNA-sequencing datasets has identified an SMG-enriched gene expression signature comprising 289 protein-coding and 75 long non-coding RNA (lncRNA) genes that include both known regulators of salivary gland function and several novel candidates ripe for future exploration. To complement these transcriptomic studies, we have generated chromatin immunoprecipitation sequencing (ChIP-seq) datasets of key histone modifications on human SMGs. Our epigenomic analyses have allowed us to identify genome-wide enhancers and super-enhancers that are likely to drive genes and regulatory pathways that are important in human SMG biology. Finally, comparative analysis with mouse and human SMG and other tissue datasets reveals evolutionary conserved gene and regulatory networks, underscoring fundamental mechanisms of salivary gland biology. Collectively, this study offers a valuable knowledge-based resource that can facilitate targeted research on salivary gland dysfunction in human patients. Full article

Low-altitude unmanned aerial vehicle (UAV) technology serves as a crucial pathway for developing a low-carbon circular economy and achieving the Sustainable Development Goals (SDGs). In order to achieve the diffusion of low-altitude UAV technology in sustainable development, a dynamic model of technology adoption decisions within enterprise clusters is constructed. This model is based on complex network evolutionary game theory. The present study investigates the mechanism through which government policies influence the diffusion of low-altitude UAV technology. The research findings indicate that government subsidy mechanisms and diffusion constraints play critical roles in the diffusion process of low-altitude UAV technology. Core Enterprises and Marginal Enterprises within clusters exhibit different responsiveness to subsidies, with Core Enterprises demonstrating higher sensitivity. The intensity of government subsidies is positively correlated with the diffusion rate of low-altitude UAV technology, while the penalty from constraints is negatively correlated with the diffusion rate. These findings establish a foundation for governments to devise pertinent subsidy mechanisms, establish and enhance the management system of the low-altitude economy, and cultivate a policy ecosystem conducive to the diffusion of low-altitude UAV technology, thereby propelling sustainable societal development. Full article