Search Results (1,178)

Efficient resource allocation is critical for transportation hub operations, yet current scheduling systems require substantial domain-specific customization when deployed across different facilities. This paper presents a domain-adaptive deep reinforcement learning (DADRL) framework that learns transferable optimization policies for dynamic resource allocation across structurally similar transportation scheduling problems. The framework integrates dual-level heterogeneous graph attention networks for separating constraint topology from domain-specific features, hypergraph-based constraint modeling for capturing high-order dependencies, and hierarchical policy decomposition that reduces computational complexity from $O\left(mnT\right)$ to $O(m+n+T)$. Evaluated on realistic simulators modeling airport gate assignment (Singapore Changi: 50 gates, 300–400 daily flights) and seaport berth allocation (Singapore Port: 40 berths, 80–120 daily vessels), DADRL achieves 87.3% resource utilization in airport operations and 86.3% in port operations, outperforming commercial solvers under strict real-time constraints (Gurobi-MIP with 300 s time limit: 85.1%) while operating 270 times faster (1.1 s versus 298 s per instance). Given unlimited time, Gurobi achieves provably optimal solutions, but DADRL reaches 98.7% of this optimum in 1.1 s, making it suitable for time-critical operational scenarios where exact solvers are computationally infeasible. Critically, policies trained exclusively on airport scenarios retain 92.4% performance when applied to ports without retraining, requiring only 800 adaptation steps compared to 13,200 for domain-specific training. The framework maintains 86.2% performance under operational disruptions and scales to problems three times larger than training instances with only 7% degradation. These results demonstrate that learned optimization principles can generalize across transportation scheduling problems sharing common constraint structures, enabling rapid deployment of AI-based scheduling systems across multi-modal transportation networks with minimal customization and reduced implementation costs. Full article

With the rapid development of Massive Open Online Courses (MOOCs), high dropout rates have become a major challenge, limiting the quality of online education and the effectiveness of targeted interventions. Although existing MOOC dropout prediction methods have incorporated deep learning and attention mechanisms to improve predictive performance to some extent, they still face limitations in modeling differences in course difficulty and learning engagement, capturing multi-scale temporal learning behaviors, and controlling model complexity. To address these issues, this paper proposes a MOOC dropout prediction model that integrates multi-scale convolution with a symmetric dual-attention mechanism, termed SDA-Net. In the feature modeling stage, the model constructs a time allocation ratio matrix (MRatio), a resource utilization ratio matrix (SRatio), and a relative group-level ranking matrix (Rank) to characterize learners’ behavioral differences in terms of time investment, resource usage structure, and relative performance, thereby mitigating the impact of course difficulty and individual effort disparities on prediction outcomes. Structurally, SDA-Net extracts learning behavior features at different temporal scales through multi-scale convolution and incorporates a symmetric dual-attention mechanism composed of spatial and channel attention to adaptively focus on information highly correlated with dropout risk, enhancing feature representation while maintaining a relatively lightweight architecture. Experimental results on the KDD Cup 2015 and XuetangX public datasets demonstrate that SDA-Net achieves more competitive performance than traditional machine learning methods, mainstream deep learning models, and attention-based approaches on major evaluation metrics; in particular, it attains an accuracy of 93.7% on the KDD Cup 2015 dataset and achieves an absolute improvement of 0.2 percentage points in Accuracy and 0.4 percentage points in F1-Score on the XuetangX dataset, confirming that the proposed model effectively balances predictive performance and model complexity. Full article

Accurate Short-Term Load Forecasting (STLF) is paramount for the stable and economical operation of power systems, particularly in the context of high renewable energy penetration, which exacerbates load volatility and non-stationarity. The prevailing advanced “decomposition–ensemble” paradigm, however, faces two significant challenges when processing non-stationary signals: (1) The performance of Variational Mode Decomposition (VMD) is highly dependent on its hyperparameters (K, α), and traditional meta-heuristic algorithms (e.g., GA, GWO, PSO) are prone to converging to local optima during the optimization process; (2) Deep learning predictors struggle to dynamically weigh the importance of multi-dimensional, heterogeneous features (such as the decomposed Intrinsic Mode Functions (IMFs) and external climatic factors). To address these issues, this paper proposes a novel, adaptive hybrid forecasting framework, namely IRIME-VMD-CDA-LSTNet. Firstly, an Improved Rime Optimization Algorithm (IRIME) integrated with a Gaussian Mutation strategy is proposed. This algorithm adaptively optimizes the VMD hyperparameters by targeting the minimization of average sample entropy, enabling it to effectively escape local optima. Secondly, the optimally decomposed IMFs are combined with climatic features to construct a multi-dimensional information matrix. Finally, this matrix is fed into an innovative Cross-Dimensional Attention (CDA) LSTNet model, which dynamically allocates weights to each feature dimension. Ablation experiments conducted on a real-world dataset from a distribution substation demonstrate that, compared to GA-VMD, GWO-VMD, and PSO-VMD, the proposed IRIME-VMD method achieves a reduction in Root Mean Square Error (RMSE) of up to 18.9%. More importantly, the proposed model effectively mitigates the “prediction lag” phenomenon commonly observed in baseline models, especially during peak load periods. This framework provides a robust and high-accuracy solution for non-stationary load forecasting, holding significant practical value for the operation of modern power systems. Full article

In the ecosystem, birds are important indicators that can sensitively reflect changes in the ecological environment and its health. However, bird monitoring has challenges due to species diversity, variable behaviors, and distinct morphological characteristics. Therefore, we propose a parallel dual-branch hybrid CNN–Transformer architecture for feature extraction that simultaneously captures local and global image features to address the “local feature similarity” issue in dual tasks of bird species and behaviors. The dual-task framework comprises three main components: the Token Re-segmentation Module (TRM), the Multi-scale Adaptive Module (MAM), and the Feature Interleaving Structure (FIS). The designed MAM fuses hybrid attention to address the problem of different-scale birds. MAM models the interdependencies between spatial and channel dimensions of features from different scales. It enables the model to adaptively choose scale-specific feature representations, accommodating inputs of different scales. In addition, we designed an efficient feature-sharing mechanism, called FIS, between parallel CNN branches. FIS interleaving delivers and fuses CNN feature maps across parallel layers, combining them with the features of the corresponding Transformer layer to share local and global information at different depths and promote deep feature fusion across parallel networks. Finally, we designed the TRM to address the challenge of visually similar but distinct bird species and of similar poses with distinct behaviors. TRM adopts a two-step approach: first, it locates discriminative regions, and then performs fine segmentation on them. This module enables the network to allocate relatively more attention to key areas while merging non-essential information and reducing interference from irrelevant details. Experiments on the self-made dataset demonstrate that, compared with state-of-the-art classification networks, the proposed network achieves the best performance, achieving 79.70% accuracy in bird species recognition, 76.21% in behavior recognition, and the best performance in dual-task recognition. Full article

In the current context of climate change, special attention should be paid to assuring the security of food and fruits. Lemon trees struggle to keep their physiological traits stable in the context of all the cumulated challenges originating from climate stress. Therefore, our aim was to assess two seed priming methods’ long-term effects on some physiological parameters of young lemon trees. The relative chlorophyll content reveals that hydropriming shows 26% increases from E1 to E6, similar to the control, while osmopriming has a 31% higher value at the beginning and after three years. Leaf stomatal density has 80% lower values due to osmopriming compared to the control, while hydropriming show 15% lower values. Leaf area development was slightly similar between treatments, with more leaves being developed after hydropriming treatments. Guard cell width has similar values for priming, with both being with 40% higher than that of the control. Lemon trees grown after osmotic stress have the highest mass percentages of magnesium and potassium in the leaves. Hydropriming promotes calcium oxalate accumulation and a high mass percentage of phosphorus. The percentage allocation of carbon as dry matter is 32% for osmopriming, significantly higher than for the other treatments. The quantum yield of photosynthetic electron transport is the only significant photosynthetic parameter for osmoprimed lemon young trees. Physiological techniques successfully enhanced the overall growth of three-year-old lemon trees, especially osmopriming treatment. Full article

Tutoring has played a significant role in pandemic-related learning recovery, supporting student learning and engagement. This paper follows up on a recent randomized controlled trial (RCT) estimating that one-on-one virtual early literacy tutoring was nearly twice as effective as two-on-one tutoring for improving student learning. To better understand this gap, we analyze transcripts from 16,629 tutoring sessions from this RCT—which included over 3.7 million tutor utterances—using natural language processing and machine learning techniques. We explore how tutors allocate attention across content instruction, relationship building, and classroom management between one-on-one and two-on-one formats. While tutors dedicate similar time to content instruction and relationship building across both formats, students receiving one-on-one tutoring receive more attention and personalized support. To improve the effectiveness of two-on-one tutoring, it may be beneficial to equip tutors with strategies that engage multiple students simultaneously, thereby reducing downtime and minimizing the potential for disengagement. Full article

Multi-Robot Task Allocation (MRTA) with spatiotemporal constraints presents significant challenges in environmental adaptability. Existing learning-based methods often overlook environmental spatial constraints, leading to spatial information distortion. To address this, we formulate the problem as an asynchronous Markov Decision Process over a directed heterogeneous graph and propose a novel heterogeneous graph neural network named the Edge-Enhanced Attention Network (E2AN). This network integrates a specialized encoder, the Edge-Enhanced Heterogeneous Graph Attention Network (E2HGAT), with an attention-based decoder. By incorporating edge attributes to effectively characterize path costs under spatial constraints, E2HGAT corrects spatial distortion. Furthermore, our approach supports flexible extension to diverse payload scenarios via node attribute adaptation. Extensive experiments conducted in simulated environments with obstructed maps demonstrate that the proposed method outperforms baseline algorithms in task success rate. Remarkably, the model maintains its advantages in generalization tests on unseen maps as well as in scalability tests across varying problem sizes. Ablation studies further validate the critical role of the proposed encoder in capturing spatiotemporal dependencies. Additionally, real-time performance analysis confirms the method’s feasibility for online deployment. Overall, this study offers an effective solution for MRTA problems with complex constraints. Full article

We study conflicts between envy-based fairness and efficiency for allocating indivisible items under additive utilities. We formalize several small, transparent instances showing that standard envy-freeness (EF) or its relaxations EFX and EFX₀—i.e., envy-freeness up to any item, where EFX restricts attention to positively valued items and EFX₀ allows removing zero-valued items as well—can conflict with Pareto-optimality (PO), maximin (MM), or maximum Nash welfare (MNW). Normatively, we argue that envy-freeness (even as EFX or EFX₀) is not a panacea for allocating indivisible items and should be weighed against efficiency and welfare criteria. Full article

Attention can be automatically captured by the distractor that matches the representation of working memory (WM) in search tasks, impairing visual search efficiency and resulting in attentional capture effects. The resource hypothesis of visual search predicts that resource allocation affects attentional capture. However, previous studies have shown partly opposing results inconsistent with this prediction. The purpose of this study is to assess the connection between attentional capture and WM resource allocation. Two experiments were conducted to combine the attentional capture paradigm with continuous delayed-estimation tasks. In Experiment 1, we manipulated the number of memory items between one and two and measured the WM representation precision as well as the magnitude of attentional capture. In Experiment 2, we manipulated resource allocation using a retro-cue task with the presentation of two memory items. In Experiment 1, the results show that when remembering one item, a single-item representation had higher precision compared to the scenario for remembering two items, and it also involved a greater allocation of WM resources. However, there was no significant difference in the magnitude of attentional capture effects between the two conditions. In Experiment 2, the results show that memory precision was higher when the cue pointed to the item compared to when it did not, but there was no significant difference in the magnitude of attentional capture effects between the cued-match and non-cued-match conditions. The findings show that the size of attentional capture effects based on WM is unaffected by the distribution of WM resources. Attentional capture effects may reflect the attention bias of WM representation that occurs in preparation stage of memory-based attentional guidance. Full article

Satellite Clock Bias (SCB) is a major source of error in Precise Point Positioning (PPP). The real-time service products from the International GNSS Service (IGS) are susceptible to network interruptions. Such disruptions can compromise product availability and, consequently, degrade positioning accuracy. We introduce the CNN-LSTM-Attention model to address this challenge. The model enhances a Long Short-Term Memory (LSTM) network by integrating Convolutional Neural Networks (CNNs) and an Attention mechanism. The proposed model can efficiently extract data features and balance the weight allocation in the Attention mechanism, thereby improving both the accuracy and stability of predictions. Across various forecasting horizons (1, 2, 4, and 6 h), the CNN-LSTM-Attention model demonstrates prediction accuracy improvements of (76.95%, 66.84%, 65.92%, 84.33%, and 43.87%), (72.59%, 65.61%, 74.60%, 82.98%, and 51.13%), (70.45%, 68.52%, 81.63%, 88.44%, and 60.49%), and (70.26%, 70.51%, 84.28%, 93.66%, and 66.76%), respectively, across the five benchmark models: Linear Polynomial (LP), Quadratic Polynomial (QP), Autoregressive Integrated Moving Average (ARIMA), Backpropagation Neural Network (BP), and LSTM models. Furthermore, in dynamic PPP experiments utilizing IGS tracking stations, the model predictions achieve positioning accuracy comparable to that of post-processed products. This proves that the proposed model demonstrates superior accuracy and stability for predicting SCB, while also satisfying the demands of positioning applications. Full article

Emergency supplies policies are a key component of regional risk governance, yet their design coherence has received limited systematic examination. Focusing on the Yangtze River Delta (YRD), this study conducts a design-oriented evaluation of emergency supplies policy design by integrating policy text mining with the Policy Modeling Consistency (PMC) index model. Based on a corpus of 212 emergency supplies–related policy documents, the study first examines the structural features and thematic emphases of the regional policy system and constructs a PMC-based evaluation framework within a mission–structure–mechanism perspective. On this basis, 16 provincial- and municipal-level policies issued between 2019 and 2023 are identified as core, system-defining policy texts and subjected to in-depth PMC evaluation. The results indicate that the evaluated core emergency supplies policies exhibit an overall “good” level of design coherence. Mission-oriented dimensions, including normative orientation and policy objectives, are generally well articulated, whereas mechanism-oriented dimensions—particularly linkage response and allocation arrangements—are specified less consistently. Observed interjurisdictional differences reflect institutional roles and governance traditions rather than variations in administrative capacity. By shifting analytical attention from implementation outcomes to design-stage coherence in core policy texts, this study provides a structured diagnostic approach for assessing emergency supplies policy design and offers insights for strengthening regional coordination and institutional resilience. Full article

With the rise in experience economy and the popularization of digital technology, user-generated content (UGC) has become a core data source for understanding tourist needs and evaluating the service quality of venues. As a landmark venue that combines science education, interactive experience, and landscape viewing, the service quality of the Macau Science Center directly affects tourists’ travel experience and word-of-mouth dissemination. However, existing studies mostly rely on traditional questionnaire surveys and lack multi-technology collaborative analysis. In order to accurately identify the factors affecting satisfaction, this study uses 788 valid UGC data from five major platforms, namely Google Maps reviews, TripAdvisor, Sina Weibo, Xiaohongshu (Rednote), and Ctrip, from January 2023 to November 2025. It integrates word frequency analysis, semantic network analysis, latent Dirichlet allocation (LDA) topic modeling, and Valence Aware Dictionary and sEntiment Reasoner (VADER) sentiment computing to construct a systematic research framework. The study found that (1) the core attention dimensions of users cover the needs of parent–child and family visits, exhibitions and interactive experiences, ticketing and consumption services, surrounding environment and landscape, emotional evaluation, and recommendation intention. (2) The keyword association network has gradually developed from a loose network in the early stage to a comprehensive experience-dense network. (3) LDA analysis identified five main potential demand themes: comprehensive visiting experience and scenario integration, parent–child interaction and characteristic scenario experience, core venue facilities and ticketing services, visiting value and emotional evaluation, and transportation and surrounding landscapes. (4) User emotions were predominantly positive, accounting for 82.7%, while negative emotions were concentrated in local service details, and the emotional scores showed a fluctuating upward trend. This study provides targeted suggestions for the service optimization of the Macau Science Center and also provides a methodological reference for UGC-driven research in similar cultural venues. Full article

Vehicle-to-Everything (V2X) has become a key technology for addressing intelligent transportation challenges. Improving spectrum utilization and mitigating multi-user interference among V2X links are currently the primary focuses of research efforts. However, the time-varying nature of channel resources and the dynamic vehicular environment pose significant challenges to achieving high spectral efficiency and low interference. Numerous studies have demonstrated the effectiveness of deep reinforcement learning (DRL) in distributed resource allocation for vehicular networks. Nevertheless, in conventional distributed DRL frameworks, the independence of agent decisions often weakens cooperation among agents, thereby limiting the overall performance potential of the algorithms. To address this limitation, this paper proposes a state-aware communication resource allocation algorithm for vehicular networks. The proposed approach enhances the representation capability of observable data by expanding the state space, thus improving the utilization of available observations. Additionally, a conditional attention mechanism is introduced to strengthen the model’s perception of environmental dynamics. These innovative improvements significantly enhance each agent’s awareness of the environment and promote effective collaboration among agents. Simulation results verify that the proposed algorithm effectively improves agents’ environmental perception and inter-agent cooperation, leading to superior performance in complex and dynamic V2X scenarios. Full article

Within the context of China’s green economy aimed at sustainable development, research on the linkage between water resources and industry has garnered considerable attention in the academic community. However, the impact of total body water (TBW) transfer and allocation embodied in the labor force—the primary economic actors—has not been addressed in the economic sector. On methodology, the “EEIO-SDA-SPD-II” (ISSI) model employed in this study encompasses measurements methods, such as an environmentally extended input–output model (EEIO), structural decomposition analysis (SDA), structural path decomposition (SPD), and the imbalance index (II), to explore the crucial paths, driving factors, and distribution of water transfer in TWB spanning 15 Chinese industries between 2007 and 2022. The findings indicate that the shifts in TBW in the manufacturing sector are more discernible when viewed through the lens of social driving factors. The construction business exhibits the most significant increase in male total body water (MTBW), whereas the education sector reflects the rapid growth in female total body water (FTBW). Pertaining to final demand, domestic consumption constitutes the primary contributor category to the increase in TWB, followed by fixed capital formation and exports. According to the SPD results, the construction sector exerts the greatest influence on the transfer of MTBW, while the education sector is characterized by the highest path coefficient value for FTBW. In contrast, the manufacturing sector shows the most pronounced initial path. Based on the imbalance index analysis, agriculture derives the greatest economic gains from TBW input, whereas the education sector yields the lowest. Full article

Forest ecological compensation policies (FECPs) are a key institutional arrangement for balancing ecological conservation and farmers’ development needs in national parks. Existing research has often treated such policies as a homogeneous whole, failing to clearly reveal the mechanisms through which different policy types affect farmers’ livelihoods, while also paying insufficient attention to complex property-rights settings. This study takes Wuyi Mountain National Park—a typical representative of collective forest regions in southern China—as a case study. Based on 239 micro-survey datasets from farming households and employing the mprobit model and moderating effect models, it investigates the influence, mechanisms, and heterogeneity of farmers’ livelihood capital in terms of their livelihood strategy choices under the moderating roles of “blood-transfusion” and “blood-making” FECPs. The results show the following: (1) Among the sample farmers, livelihood strategies are distributed as follows: pure agricultural type (31.8%), out-migration for work type (20.5%), and commercial operation type (47.7%). (2) Farmers’ livelihood capital has a significant impact on their livelihood strategy choice, with different dimensions of capital playing distinct roles. (3) FECPs follow differentiated moderating pathways. “Blood-transfusion” policies emphasize compensation and buffering functions, reducing farmers’ livelihood transition pressure through direct cash transfers; “blood-making” policies reflect empowerment and restructuring characteristics, activating physical assets and reshaping the role of social capital through productive investment. Together, they constitute a complementary system of protective security and transformative empowerment. Accordingly, this study proposes policy insights such as building a targeted ecological compensation system that is categorized, dynamically linked, and precise; innovating compensation fund allocation mechanisms that integrate collective coordination with household-level benefits; optimizing policy design oriented toward enhancing productive capital; and establishing robust monitoring, evaluation, and adaptive management mechanisms for dynamic FECPs. Full article