Spatial Weight Research Articles

Temporal-spatial dependencies enhanced deep learning model for time series forecast

Forecasting financial time series can be challenging because of the inherent complex interplay of temporal and spatial dynamics in the data. Specifically, time series for different geographical locations often have different yet inter-dependent patterns; even for the same-location time series, patterns can vary over time. Therefore, an accurate forecasting model has to take both temporal and spatial patterns into consideration. This study proposes a new model named Temporal-Spatial dependencies ENhanced deep learning model (TSEN), which consists of multiple RNN-based layers and an attention layer: each RNN-based layer learns the temporal pattern of a specific time series at the presence of multivariate exogenous series, and the attention layer learns the spatial correlative weight and obtains the global representations simultaneously. We illustrate the use of TSEN in forecasting household leverage time series in China. We show that clustering and choosing correlative series are necessary steps to obtain accurate forecast. The results show that the proposed method captures the temporal-spatial dynamics of household leverage and generates more accurate prediction than traditional methods.

How do innovation factor allocation and institutional environment affect high-quality economic development? Evidence from China

An in-depth understanding of the relationship between innovation factor allocation, the institutional environment, and high-quality economic development can help us better comprehend how better productivity promotes high-quality, regionally coordinate, and balanced economic development. Using panel data of 30 provincial-level regions in China from 2011 to 2019, we explore the effect of innovation factor allocation on high-quality economic development and the moderating effects of different institutional environments. First, we construct a comprehensive indicator system for innovation factor allocation and a new model to calculate the weights of various indicators of innovation factors. Second, we apply TOPSIS and grey correlation to measure the level of innovation factor allocation, and construct a modified gravity model to generate a directed weighted spatial association network of this allocation as a spatial weight matrix. Finally, we use the spatial Durbin model to empirically test the spatial effects of innovation factor allocation on high-quality economic development under different institutional environments. We find that: First, innovation factor allocation significantly positively affects the high-quality development of the local economy, while it has negative spatial spillover effects on neighboring areas. The impact of different types of institutional environments on high-quality economic development exhibits an inverted “U” shape. China's institutional environment has not crossed the critical threshold of this “U" shaped relationship. Thus, different types of institutional environments can still effectively strengthen the role of innovation factor allocation in promoting high-quality economic development. In particular, the strengthening effect of online government is the strongest and its negative impact is the weakest. Overall, this study is of great significance for optimizing innovation factor allocation in various regions and promoting high-quality economic development.

Spatial Spillover of the Global Internet Penetration Rate and the Digital Gender Divide

This study investigated the possible contributing factors to the digital gender divide and the spatial spillover effects of submarine communication cables on neighboring areas. Using spatial statistical models and global data, this study examined the Internet penetration rate and the digital gender divide in 186 countries. Data sources included open source data, social media, ITU and UNCTAD databases, and the World Bank. The study estimated the average male and female Internet users in 186 countries using Facebook API. The number of submarine communication cables between countries was used as the basis for constructing the spatial weights matrix. The countries with the highest Internet penetration rate were mostly in Europe. Those with the lowest Internet penetration rate were in Africa, Central Asia, and West Asia. Thirteen of the 15 countries with the worst Internet penetration rate were located in the African region. A significant digital gender divide arose in the developing and the least-developed countries. We argue that connectivity is key to the reduction of the digital gender divide and to ensuring that more women have access to the Internet. The study ascertains that the Internet penetration rate is related to the digital gender divide in individual countries and has a spatial spillover impact in surrounding nations connected by submarine communication cables. The study demonstrates that submarine cables serve as important infrastructure for closing the digital gender divide and deserve increased analytical attention.

The distance decay effect and spatial reach of spillovers

This paper quantifies and graphically illustrates the distance decay effect and spatial reach of spillover effects derived from a spatial Durbin (SD) model with parameterized spatial weight matrices. Building on attributes of the concept of spatial autocorrelation developed by Arthur Getis, we adopt a distance-based negative exponential spatial weight matrix and parameterize it by a decay parameter that is different for each spatial lag in this model, both of the regressand and of all regressors. The quantification and illustration are applied to the spatially augmented neoclassical growth framework, which we estimate using data for 266 NUTS-2 regions in the EU over the period 2000–2018. We find distance decay parameters ranging from 0.233 to 2.224 and spatial reaches ranging from 700 to more than 1500 km for the different growth determinants in this model. These wide ranges highlight the restrictiveness of the conventional SD model based on one common spatial weight matrix for all spatial lags.

New Progress in Intelligent Picking: Online Detection of Apple Maturity and Fruit Diameter Based on Machine Vision

In the realm of automated apple picking operations, the real-time monitoring of apple maturity and diameter characteristics is of paramount importance. Given the constraints associated with feature detection of apples in automated harvesting, this study proposes a machine vision-based methodology for the accurate identification of Fuji apples’ maturity and diameter. Firstly, maturity level detection employed an improved YOLOv5s object detection model. The feature fusion section of the YOLOv5s network was optimized by introducing the cross-level partial network module VoVGSCSP and lightweight convolution GSConv. This optimization aimed to improve the model’s multiscale feature information fusion ability while accelerating inference speed and reducing parameter count. Within the enhanced feature fusion network, a dual attention mechanism combining channel and spatial attention (GAM) was introduced to refine the color and texture feature information of apples and to increase spatial position feature weights. In terms of diameter determination, the contours of apples are obtained by integrating the dual features of color and depth images within the target boxes acquired using the maturity detection model. Subsequently, the actual area of the apple contour is determined by calculating the conversion relationship between pixel area and real area at the current depth value, thereby obtaining the diameter of the apples. Experimental results showed that the improved YOLOv5s model achieved an average maturity level detection precision of 98.7%. Particularly noteworthy was the detection accuracy for low maturity apples, reaching 97.4%, surpassing Faster R-CNN, Mask R-CNN, YOLOv7, and YOLOv5s models by 6.6%, 5.5%, 10.1%, and 11.0% with a real-time detection frame rate of 155 FPS. Diameter detection achieved a success rate of 93.3% with a real-time detection frame rate of 56 FPS and an average diameter deviation of 0.878 mm for 10 apple targets across three trials. Finally, the proposed method achieved an average precision of 98.7% for online detection of apple maturity level and 93.3% for fruit diameter features. The overall real-time inference speed was approximately 56 frames per second. These findings indicated that the method met the requirements of real-time mechanical harvesting operations, offering practical importance for the advancement of the apple industry.

3D suitability evaluation of urban underground space using a variable weight method and considering ground restrictions

The evaluation of urban underground space (UUS) suitability involves multiple indicators. Assigning weight to these indicators is crucial for accurate assessment. This paper presents a method for spatially variable weight assignment of indicators using the order relation analysis method (G1-method), the entropy weight method, an improved grey relational analysis (GRA) and a set of spatial weight adjustment coefficients. First, the subjective and objective weights of indicators for engineering geological and hydrogeological conditions were determined by the G1-method and entropy weight method, respectively, and their combined weights were then obtained using the principle of minimum discriminatory information. This study highlighted the impact of surface restrictions, such as buildings, on UUS, and the degree of the influence of these buildings gradually decreased with the increase in depth of the rock and soil mass in UUS, which resulted in changes in weights of indicators with depth. To address this issue, a coefficient was defined as the standardized value of the ratio of additional stress applied by restrictions to the self-weight stress of soil at the same depth to modify the combined weights so that all weights of indicators could vary in space. Finally, an improved GRA was used to determine the suitability level of each evaluation cell using the maximum correlation criterion. This method was applied to the 3D suitability evaluation of UUS in Sanlong Bay, Foshan City, Guangdong Province, China, including 16 evaluation indexes. This study comprehensively considered the influence of multiple factors, thereby providing reference for evaluating the suitability of UUS in big cities.

S2WAT: Image Style Transfer via Hierarchical Vision Transformer Using Strips Window Attention

Transformer's recent integration into style transfer leverages its proficiency in establishing long-range dependencies, albeit at the expense of attenuated local modeling. This paper introduces Strips Window Attention Transformer (S2WAT), a novel hierarchical vision transformer designed for style transfer. S2WAT employs attention computation in diverse window shapes to capture both short- and long-range dependencies. The merged dependencies utilize the "Attn Merge" strategy, which adaptively determines spatial weights based on their relevance to the target. Extensive experiments on representative datasets show the proposed method's effectiveness compared to state-of-the-art (SOTA) transformer-based and other approaches. The code and pre-trained models are available at https://github.com/AlienZhang1996/S2WAT.

High-Performance Defect Detection Methods for Real-Time Monitoring of Ceramic Additive Manufacturing Process Based on Small-Scale Datasets

Vat photopolymerization is renowned for its high flexibility, efficiency, and precision in ceramic additive manufacturing. However, due to the impact of random defects during the recoating process, ensuring the yield of finished products is challenging. At present, the industry mainly relies on manual visual inspection to detect defects; this is an inefficient method. To address this limitation, this paper presents a method for ceramic vat photopolymerization defect detection based on a deep learning framework. The framework innovatively adopts a dual-branch object detection approach, where one branch utilizes a fully convolution network to extract the features from fused images and the other branch employs a differential Siamese network to extract the differential information between two consecutive layer images. Through the design of the dual branches, the decoupling of image feature layers and image spatial attention weights is achieved, thereby alleviating the impact of a few abnormal points on training results and playing a crucial role in stabilizing the training process, which is suitable for training on small-scale datasets. Comparative experiments are implemented and the results show that using a Resnet50 backbone for feature extraction and a HED network for the differential Siamese network module yields the best detection performance, with an obtained F1 score of 0.89. Additionally, as a single-stage defect object detector, the model achieves a detection frame rate of 54.01 frames per second, which meets the real-time detection requirements. By monitoring the recoating process in real-time, the manufacturing fluency of industrial equipment can be effectively enhanced, contributing to the improvement of the yield of ceramic additive manufacturing products.

Study of land potential based on soil characteristics using interpolation and scoring methods for non agricultural uses: a case study in Wonosalam, Jombang District, East Java

The state asset land managed by UPN "Veteran" East Java is located in the Wonosalam District based on land certificate data issued by the Jombang BPN covering 13.5 hectares. Currently more than 50% of the land has not been used optimally to support learning, research and community service. The problems that arise in utilizing land assets to support learning, research and community service are 1) the legal boundaries of land ownership are not yet clear, 2) there is no land use master plan to support the learning process, 3) there is no potential data that can be used. prepare use and development plans. Against the background of the problems above, this research aims to map the potential of state land assets by "spatial interpolation, scoring and weight overlay" of land characteristics. Interpolation is a way of calculating the area of several representative measurement points, while scoring and "Weight Overlay" is a facility provided by ArcMap GIS which combines input spatial data with weighting (weight factor) and scoring. Land unit maps are obtained from overlaying relief, geology, slope and land use maps. Interpolation, Scoring and Weight Overlay of the nine land characteristics yielded a potential land asset for buildings, good category 16.61% (44,130.38 m2); moderate 25.66% (67,382.75 m2) and bad 7.523% (19,784.88 m2). Playing field potential, good category 54.73% (71,865.05 m2) and medium 45.27% (59,432.62 m2). Potential for road infrastructure development in the good category 32.38% (42513.74 m2); Moderate 42.68% (56035.24 m2) and Bad 24.94% (32748.69 m2), and all of these potentials are in Slightly Flat topography (Slope, 0-3%) to hilly (Slope, 15-25%).

Flow Spatiotemporal Moran's I: Measuring the Spatiotemporal Autocorrelation of Flow Data

Flows can reflect the spatiotemporal interactions or movements of geographical objects between different locations. Measuring the spatiotemporal autocorrelation of flows can help determine the overall spatiotemporal trends and local patterns. However, quantitative indicators of flows used to measure spatiotemporal autocorrelation both globally and locally are still rare. Therefore, we propose the global and local flow spatiotemporal Moran's I (FSTI). The global FSTI is used to assess the overall spatiotemporal autocorrelation degree of flows, and the local FSTI is applied to identify local spatiotemporal clusters and outliers. In the FSTI, to reflect flow spatiotemporal adjacency relationships, we establish flow spatiotemporal weights by multiplying the spatial and temporal weights of flows considering spatiotemporal orthogonality. The flow spatial weights include contiguity‐based (considering first/higher‐order and common border) and Euclidean distance‐based weights. The temporal weights consider ordinary and lagged cases. As flow attributes may follow a long‐tail distribution, we conduct Monte Carlo simulations to evaluate the statistical significance of the results. We assess the FSTI using synthetic datasets and Chinese population mobility datasets, and compare some results with those of recent flow‐related methods. Additionally, we perform a sensitivity analysis to select a suitable temporal threshold. The results show that the FSTI can be used to effectively detect spatiotemporal variations in the autocorrelation degree and type.

RAAWC-UNet: an apple leaf and disease segmentation method based on residual attention and atrous spatial pyramid pooling improved UNet with weight compression loss.

Early detection of leaf diseases is necessary to control the spread of plant diseases, and one of the important steps is the segmentation of leaf and disease images. The uneven light and leaf overlap in complex situations make segmentation of leaves and diseases quite difficult. Moreover, the significant differences in ratios of leaf and disease pixels results in a challenge in identifying diseases. To solve the above issues, the residual attention mechanism combined with atrous spatial pyramid pooling and weight compression loss of UNet is proposed, which is named RAAWC-UNet. Firstly, weights compression loss is a method that introduces a modulation factor in front of the cross-entropy loss, aiming at solving the problem of the imbalance between foreground and background pixels. Secondly, the residual network and the convolutional block attention module are combined to form Res_CBAM. It can accurately localize pixels at the edge of the disease and alleviate the vanishing of gradient and semantic information from downsampling. Finally, in the last layer of downsampling, the atrous spatial pyramid pooling is used instead of two convolutions to solve the problem of insufficient spatial context information. The experimental results show that the proposed RAAWC-UNet increases the intersection over union in leaf and disease segmentation by 1.91% and 5.61%, and the pixel accuracy of disease by 4.65% compared with UNet. The effectiveness of the proposed method was further verified by the better results in comparison with deep learning methods with similar network architectures.

Access Weight Matrix: A Place and Mobility Infused Spatial Weight Matrix

This study introduces the Access Weight Matrix (AWM) to capture the spatial dependence of access across a geographical surface. AWM is a nonsymmetry, nonzero diagonal matrix with elements to be a function of (i) the spatial distribution of places, (ii) the number of places, and (iii) the travel‐time threshold to reach places rather than distance, contiguity, or adjacency. AWM is tested and validated to examine the spatial dependence of transit access to employment opportunities in the City of Chicago. Three observations are noticed. First, the degree of spatial dependence between the access of geographical units is not necessarily proportional to their proximity and is better explained by AWM than traditional spatial weight matrices regardless of the travel‐time threshold. Second, the time‐dependence feature of AWM improves the accuracy of capturing spatial dependence, particularly in short travel‐time thresholds. Third, near geographical units are not necessarily more related than distant geographical units even for access that is proved to be spatially highly correlated with neighboring units. With the increased ease of measuring access, research is expanding to explore the socioeconomic, demographic, and built‐environment correlates of access. AWM can be employed in developing more accurate spatial econometrics models.

Optimized ST-Moran’s I Model for Characterizing the Dynamic Evolution of Terminal Airspace Congestion

This study aims to unveil the spatiotemporal evolution of congestion within terminal airspace, offering an in-depth analysis of congestion concerns to effectively utilize airspace resources and devise targeted control strategies, thereby enhancing airspace operation safety and efficiency. Initially, converting segment flow rates into equivalent speeds serves as a quantitative benchmark for operational status. Subsequently, an enhanced version of the ST-Moran’s I index model, specifically tailored to terminal airspace, is developed by incorporating improvements across spatial weight matrices, standard state parameters, and temporal dimensions. Validating this model with actual operational data from Chengdu’s terminal airspace, the research demonstrates significant advancements. Compared to conventional models, the proposed model enhances recognition rates for congestion in spatial and temporal dimensions by 62.5% and 43.61%, respectively. Congestion within terminal airspace predominantly occurs at the intersection of departure-climb and approach-departure segments, exhibiting evident spatiotemporal migration behavior. The proposed model accurately delineates the spatiotemporal characteristics of segment congestion, offering support for tailored congestion management strategies.

A novel dynamic spatio-temporal graph convolutional network for wind speed interval prediction

It is crucial to predict the wind speed for the utilization of renewable wind energy and the operation of transmission lines with increased capacity. The intermittency and stochastic fluctuations of wind speed pose a significant challenge for the high-quality wind speed prediction, and a novel wind speed interval prediction (WSIP) model is constructed in this study by employing the residual estimation (RE)-oriented dynamic spatio-temporal graph convolutional network (DSTGCN) approach. Firstly, a dynamic adjacency matrix is designed to obtain time-varying global spatial weight allocations among each wind speed node. Then, the spatio-temporal features are extracted by using gated recurrent units (GRUs) and GCNs to construct the wind speed graph networks. Moreover, the RE-oriented strategy incorporating the pinball loss is designed to provide a guidance the parameter training of the constructed model, thus eliminating the quantile crossings problem. As a result, the deterministic point prediction of the wind speed is expanded to the quantile-based probabilistic interval prediction. Finally, the experimental results are presented to demonstrate the validity and superiority of proposed scheme in both qualitative and quantitative performance.

Can digital finance development drive green transformation in manufacturing? Evidence from China.

Digital finance is a product of emerging technology-enabled innovation in financial services and has a critical impact on the green transformation of the manufacturing industry. We propose a new efficiency measurement model based on the slacks-based measure (SBM) to measure the efficiency of green transformation on regional manufacturing. Chinese interprovincial data from 2010 to 2019 were obtained for the study. In addition, we estimated the effect of digital finance on green transformation of manufacturing using a benchmark panel model. Finally, considering the regional heterogeneity and spatial effects of green transformation efficiency in the manufacturing industry, we constructed a spatial Durbin model based on an economic-geographic nested spatial weight matrix to analyze the spatial influence of digital finance on green transformation in the manufacturing industry. The results show that the green transformation of the manufacturing industry has significant positive spatial spillover effects owing to the existence of competition, demonstration, and economic correlation effects among regions.

Estimating urban land subsidence with satellite data using a spatially multiscale geographically weighted regression approach

Land subsidence has attracted widespread attention because of its potential hazards to sustainable urban development. The traditional methods for estimating land subsidence disregard the crucial aspect of altitude by analyzing spatial data solely on the 2D plane (longitude and latitude geographic coordinates). This research incorporates the altitude feature into the construction of the spatial matrix to develop a new method named spatially multiscale geographically weighted regression (SM-GWR) for a better estimation of land subsidence, enabling to determine the optimal bandwidth at 3D spatial scales. By adopting a 3D perspective, the new algorithm accurately fits settlement points based on spatial distances, addressing previous limitations of projecting points on a two-dimensional surface and ignoring height differences, thereby providing a more realistic depiction of spatial relationships. The satellite data is first used to monitor land subsidence through the small baseline subset synthetic aperture radar interferometry (SBAS-InSAR). Then, the SM-GWR model is developed to construct the 3D spatial weight matrix and update estimate parameters through back-fitting calibration. The proposed method's feasibility is confirmed through examples of land subsidence in China and Singapore. The case study results demonstrate that: (1) The SM-GWR model shows significant improvement in fitting land subsidence compared to the traditional method (R2 value of 0.908 and 0.934 in predicting annual subsidence and cumulative subsidence, respectively); (2) The contribution of the influencing factors is highly related to the nature properties and urban development of the city. The novelty of this study lies in the development of an SM-GWR model combined with satellite monitoring to achieve high-precision estimation of land subsidence, considering the 3D structure of spatial geographic data.

Drug-target Affinity Prediction by Molecule Secondary Structure Representation Network.

Identification of drug-target interactions (DTI) is a crucial step in drug development with high specificity and low toxicity. To accelerate the process, computer-aided DTI prediction algorithms have been used to screen compounds or targets rapidly. Furthermore, DTI prediction can be used to identify potential targets for existing drugs, thus uncovering new indications and repositioning them. Therefore, it is of great importance to develop efficient and accurate DTI prediction algorithms. Current algorithms usually represent drugs as extracted features, which are learned by convolutional neural networks (CNNs) from its linear representation, or utilize graph neural networks (GNNs) to learn its graph representation. However, these methods either lose information or fail to capture the structural information of the drug. To address this issue, a novel molecule secondary structure representation network (MSSRN) is proposed to learn drug characterization more accurately. Firstly, the network performs relational graph convolutional networks (R-GCNs) on the drug's molecular graph and integrates drug sequence convolutions to learn the sequential information. Secondly, inspired by the attention mechanism, spatial importance weights of the drug sequence are calculated to guide R-GCNs to learn the topological information of the drug. A drug-target affinity model, called MSSRN-DTA, was then constructed by using MSSRN to learn drug structure and CNN to learn protein sequence. The effectiveness of the proposed method is verified by comparing it with other alternative methods and baseline models on two benchmark datasets.

Insulator defect detection based on improved Yolov5s

The stable operation of a power supply system is inseparable from the work of detecting defects in transmission lines. However, the insulator defect detection model based on deep learning is widely used in wire inspection work. Therefore, this paper proposes an improved YOLOv5s insulator defect detection model in order to solve the problems of insufficient training data and low recognition accuracy of the target detection model in the real-time detection of small target insulator defects. To expand and enhance the training data, experiments were conducted using the addition of noise and random black blocks. The spatial and channel weight coefficients were obtained by adding an attention mechanism (Convolutional Block Attention Module, CBAM), and the dimensions of the input feature maps were transformed to enhance the model’s ability to extract and fuse small target defect features. Experiments show that with Faster RCNN, YOLOv3, SSD and YOLOv4 comparison experiments verified that the algorithm achieves 97.38% detection accuracy for insulators and 93.32% detection accuracy for small target insulator defects with a fast detection speed, which is a better solution to the problem of detecting insulator defects with too small a proportion in the image.

A Methodological Approach for Gap Filling of WFV Gaofen-1 Images from Spatial Autocorrelation and Enhanced Weighting

Clouds and cloud shadow cover cause missing data in some images captured by the Gaofen-1 Wide Field of View (GF-1 WFV) cameras, limiting the extraction and analysis of the image information and further applications. Therefore, this study proposes a methodology to fill GF-1 WFV images using the spatial autocorrelation and improved weighting (SAIW) method. Specifically, the search window size is adaptively determined using Getis-Ord Gi* as a metric. The spatial and spectral weights of the pixels are computed using the Chebyshev distance and spectral angle mapper to better filter the suitable similar pixels. Each missing pixel is predicted using linear regression with similar pixels on the reference image and the corresponding similar pixel located in the non-missing region of the cloudy image. Simulation experiments showed that the average correlation coefficient of the proposed method in this study is 0.966 in heterogeneous areas, 0.983 in homogeneous farmland, and 0.948 in complex urban areas. It suggests that SAIW can reduce the spread of errors in the gap-filling process to significantly improve the accuracy of the filling results and can produce satisfactory qualitative and quantitative fill results in a wide range of typical land cover types and has extensive application potential.

OII: An Orientation Information Integrating Network for Oriented Object Detection in Remote Sensing Images

Oriented object detection for remote sensing images poses formidable challenges due to arbitrary orientation, diverse scales, and densely distributed targets (e.g., across terrain). Current investigations in remote sensing object detection have primarily focused on improving the representation of oriented bounding boxes yet have neglected the significant orientation information of targets in remote sensing contexts. Recent investigations point out that the inclusion and fusion of orientation information yields substantial benefits in training an accurate oriented object system. In this paper, we propose a simple but effective orientation information integrating (OII) network comprising two main parts: the orientation information highlighting (OIH) module and orientation feature fusion (OFF) module. The OIH module extracts orientation features from those produced by the backbone by modeling the frequency information of spatial features. Given that low-frequency components in an image capture its primary content, and high-frequency components contribute to its intricate details and edges, the transformation from the spatial domain to the frequency domain can effectively emphasize the orientation information of images. Subsequently, our OFF module employs a combination of a CNN attention mechanism and self-attention to derive weights for orientation features and original features. These derived weights are adopted to adaptively enhance the original features, resulting in integrated features that contain enriched orientation information. Given the inherent limitation of the original spatial attention weights in explicitly capturing orientation nuances, the incorporation of the introduced orientation weights serves as a pivotal tool to accentuate and delineate orientation information related to targets. Without unnecessary embellishments, our OII network achieves competitive detection accuracy on two prevalent remote sensing-oriented object detection datasets: DOTA (80.82 mAP) and HRSC2016 (98.32 mAP).