RESUMO
Coastal wetland ecosystems harbor rich biodiversity and possess significant ecosystem service value (ESV). Therefore, it offers a range of crucial ecosystem services (ES) for human well-being and socio-economic development. Taking the Hainan Island coastal zone (HICZ) as a case study, the spatio-temporal characteristics of land use and land cover change (LULCC), and its associated ESV in wetland landscapes were analyzed over three time points (2000, 2010 and 2020). We explored the spatio-temporal evolution trajectory of ESV on the basis of geo-information tupu. Then, future land use simulation (FLUS) was employed to predict wetland patterns and ESV under three different scenarios: business as usual (BAU), ecological conservation first (ECF), and economic development first (EDF). The results showed that over the past two decades, a significant proportion (exceeding 80%) of the overall wetland region was comprised of offshore and coastal wetlands (OCW) as well as constructed wetlands (CW); these formed the matrix of the landscape. The area of building land (BL) continued to exhibit a consistent upward trend. Expanding by 2.18 times, it represented the most significant increase in the rate of dynamic changes in BL. The main ES in the HICZ corresponded to the regulation services (53.57%) and the support services (27.58%). The ESV of wetland losses accounted for 45.17% (43.08 × 108 yuan) of the total loss. The spatial differentiation of ESV in the HICZ was larger in the southwest and the northeast regions, while it was comparatively lower in the north. The transformation in the area of early and late change types accounted for 236.46 km2 and 356.69 km2, respectively. The scenario ECF was achieved with an optimal development of ESV (1807.72 × 108 yuan), which was coordinated with the high-level of development of regional ES functions and the economy. These findings provide valuable information for the sustainable development as well as the protection of ecology and environment of the coastal zone under the background of the construction of Hainan pilot free trade zone in the future.
RESUMO
Plague, a highly infectious disease caused by Yersinia pestis, has killed millions of people in history and is still active in the natural foci of the world nowadays. Understanding the spatiotemporal patterns of plague outbreaks in history is critically important, as it may help facilitate the prevention and control for potential future outbreaks. This study's objective was to estimate the effect of the topography, vegetation, climate, and other environmental factors on the Y. pestis ecological niche. A maximum entropy algorithm spatially modelled plague occurrence data from 2004-2018 and the environmental variables to evaluate the contribution of the variables to the distribution of Y. pestis. Our results found that the average minimum temperature in September (-8 °C to +5 °C) and the sheep population density (250 sheep per km2) were influential in characterising the niche. The rim of Qinghai Lake showed more favourable conditions for Y. pestis presence than other areas within the study area. Identifying various factors will assist any future modelling efforts. Our suitability map identifies hotspots and will help public health officials in resource allocation in their quest to abate future plague outbreaks.
RESUMO
Electroencephalogram (EEG) signals offer invaluable insights into the complexities of emotion generation within the brain. Yet, the variability in EEG signals across individuals presents a formidable obstacle for empirical implementations. Our research addresses these challenges innovatively, focusing on the commonalities within distinct subjects' EEG data.We introduce a novel approach named Contrastive Learning Graph Convolutional Network (CLGCN). This method captures the distinctive features and crucial channel nodes related to individuals' emotional states. Specifically, CLGCN merges the dual benefits of Contrastive Learning's synchronous multisubject data learning and the Graph Convolutional Network's proficiency in deciphering brain connectivity matrices.Understanding multifaceted brain functions and their information interchange processes is realized as CLGCN generates a standardized brain network learning matrix during a dataset's learning process. Our model significantly streamlines the retraining process for new subjects, requiring only 5% of the initial sample size for fine-tuning to attain a remarkable 92.8% accuracy rate. Additionally, our model has undergone extensive testing on the DEAP and SEED datasets, demonstrating the effectiveness of our model. .
RESUMO
EEG-based emotion recognition is becoming crucial in brain-computer interfaces (BCI). Currently, most researches focus on improving accuracy, while neglecting further research on the interpretability of models, we are committed to analyzing the impact of different brain regions and signal frequency bands on emotion generation based on graph structure. Therefore, this paper proposes a method named Dual Attention Mechanism Graph Convolutional Neural Network (DAMGCN). Specifically, we utilize graph convolutional neural networks to model the brain network as a graph to extract representative spatial features. Furthermore, we employ the self-attention mechanism of the Transformer model which allocates more electrode channel weights and signal frequency band weights to important brain regions and frequency bands. The visualization of attention mechanism clearly demonstrates the weight allocation learned by DAMGCN. During the performance evaluation of our model on the DEAP, SEED, and SEED-IV datasets, we achieved the best results on the SEED dataset, showing subject-dependent experiments' accuracy of 99.42% and subject-independent experiments' accuracy of 73.21%. The results are demonstrably superior to the accuracies of most existing models in the realm of EEG-based emotion recognition.
RESUMO
Named entity recognition and relation extraction are two important fundamental tasks in natural language processing. The joint entity-relationship extraction model based on parameter sharing can effectively reduce the impact of cascading errors on model performance by performing joint learning of entities and relationships in a single model, but it still cannot essentially get rid of the influence of pipeline models and suffers from entity information redundancy and inability to recognize overlapping entities. To this end, we propose a joint extraction model based on the decomposition strategy of pointer mechanism is proposed. The joint extraction task is divided into two parts. First, identify the head entity, utilizing the positive gain effect of the head entity on tail entity identification.Then, utilize a hierarchical model to improve the accuracy of the tail entity and relationship identification. Meanwhile, we introduce a pointer model to obtain the joint features of entity boundaries and relationship types to achieve boundary-aware classification. The experimental results show that the model achieves better results on both NYT and WebNLG datasets.