Landslide susceptibility evaluation can accurately predict the spatial distribution of potential landslides, which offers great usefulness for disaster prevention, reduction, and land resource management. Aiming at problems insufficient samples landslide compilation, difficulty in expanding samples, expression nonlinear relationships among factors, this paper proposes a new method combining deep autoencoder multi-scale residual network (DAE-MRCNN). In first step, was used to learn feature...

Urban vitality serves as a crucial metric for evaluating sustainable urban development and the well-being of residents. Existing studies have predominantly focused on analyzing direct effects intensity (VI) its influencing factors, while paying less attention to diversity (VD) indirect impact mechanisms. Supported by multisource remote sensing data, this study establishes five-dimensional evaluation system employs Partial Least Squares Structural Equation Model (PLS-SEM) quantify...

Effective landslide disaster risk management contributes to sustainable development. A useful method for emergency and avoidance is Landslide Susceptibility Mapping (LSM). The statistical susceptibility prediction model based on slope unit ignores the re-lationship between triggering factors spatial characteristics. It disregards influence of adjacent image elements around slope-unit element. Therefore, this paper proposes a hardwired kernels-3DCNN approach LSMs considering spatial-factor...

To capture both global stationarity and spatiotemporal non-stationarity, a novel mixed geographically temporally weighted regression (MGTWR) model accounting for local effects in space time is presented. Since the constant spatial-temporal varying coefficients could not be estimated one step, two-stage least squares estimation introduced to calibrate model. Both simulations real-world datasets are used test verify performance of proposed MGTWR Additionally, an Akaike Information Criterion...

Geological disaster risk assessment can quantitatively assess the of disasters to hazard-bearing bodies. Visualizing geological provide scientific references for regional engineering construction, urban planning, and prevention mitigation. There are some problems in current binary classification landslide model, such as a single sample type, slow multiclass speed, large differences number positive negative samples, errors results. This paper introduces multilevel hazard scale selects...

The analysis and evaluation of landslide susceptibility are great significance in preventing managing geological hazards. Aiming at the problems insufficient information caused by limited number datasets, complex factors, low prediction accuracy susceptibility, a method based on deep attention dilated residual convolutional neural network (DADRCNN) is proposed. First, convolution unit (DCU) used to increase receptive field, aggregate multi-scale information, enhance model ability capture...

Point-of-interest (POI) recommendation is one of the fundamental tasks for location-based social networks (LBSNs). Some existing methods are mostly based on collaborative filtering (CF), Markov chain (MC) and recurrent neural network (RNN). However, it difficult to capture dynamic user’s preferences using CF methods. MC suffer from strong independence assumptions. RNN still in early stage incorporating spatiotemporal context information, main behavioral intention current sequence not...

Previous studies have demonstrated that non-Euclidean distance metrics can improve model fit in the geographically weighted regression (GWR) model. However, GWR often considers spatial nonstationarity and does not address variations local temporal issues. Therefore, this paper explores a (GTWR) approach accounts for both simultaneously to estimate house prices based on travel time metrics. Using price data collected between 1980 2016, response explanatory variables are then modeled using...

As urban development accelerates and natural disasters occur more frequently, the urgency of developing effective emergency shelter planning strategies intensifies. The location selection method under traditional multi-criteria decision-making framework suffers from issues such as strong subjectivity insufficient data support. Artificial intelligence offers a robust data-driven approach for site selection; however, many methods neglect spatial relationships targets within geographical space....

Landslide susceptibility mapping (LSM) is an important decision basis for regional landslide hazard risk management, territorial spatial planning and making. The current convolutional neural network (CNN)-based models do not adequately take into account the nature of texture features, vision transformer (ViT)-based LSM have high requirements amount training data. In this study, we overcome shortcomings CNN ViT by fusing these two deep learning (bottleneck (BoTNet) (ConViT)), fused model was...

This paper proposes an extended semi-supervised regression approach to enhance the prediction accuracy of housing prices within geographical information science field. The method, referred as co-training weighted (COGWR), aims fully utilize positive aspects both (GWR) method and learning paradigm. Housing in Beijing are assessed validate feasibility proposed model. COGWR model demonstrated a better goodness-of-fit than GWR when price data were limited because is able effectively absorb...

The quality of "non-landslide' samples data impacts the accuracy geological hazard risk assessment. This research proposed a method to improve performance support vector machine (SVM) by perfecting 'non-landslide' in landslide susceptibility evaluation model through fuzzy c-means (FCM) cluster generate more reliable maps. Firstly, three sample selection scenarios for include following principles: 1) select randomly from low-slope areas (scenario-SS), 2) with no hazards (scenario-RS), 3)...

Abstract Building shape cognition is essential for tasks, such as map generalization, urban modeling, and building semantics distribution pattern recognition. Traditional geometric statistical methods rely on human‐defined indicators, spectral‐based graph neural networks (GNNs) require Laplacian eigendecomposition, resulting in high algorithmic complexity. Therefore, we proposed a low‐complexity simple‐to‐use spatial‐domain GNN differentiating shapes. To examine the influence of vertices...

AbstractForest biomass is a significant indicator for substance accumulation and forest succession, can provide valuable information management scientific planning. Accurate estimations of at fine resolution are important better understanding the productivity carbon cycling dynamics. In this study, considering low efficiency accuracy existing estimation models remote sensing data, Landsat 8 OLI imagery field data cooperated with radial basis function artificial neural network (RBF ANN)...

With the growing popularity of location-based social media applications, point-of-interest (POI) recommendation has become important in recent years. Several techniques, especially collaborative filtering (CF), Markov chain (MC), and recurrent neural network (RNN) based methods, have been recently proposed for POI service. However, CF-based methods MC-based are ineffective to represent complicated interaction relations historical check-in sequences. Although networks (RNNs) its variants...

Geographically weighted regression (GWR) is a classical method for estimating nonstationary relationships. Notwithstanding the great potential of model processing geographic data, its large-scale application still faces challenge high computational costs. To solve this problem, we proposed computationally efficient GWR method, called K-Nearest Neighbors (KNN-GWR). First, it utilizes k-dimensional tree (KD tree) strategy to improve speed finding observations around points, and, optimize...

With the continuous accumulation of massive amounts mobile data, point-of-interest (POI) recommendation has become a vital task for location-based social networks. Deep neural networks or matrix factorization (MF) alone are challenging to effectively learn user–POI interaction functions. Moreover, is sparse, and heterogeneous characteristics auxiliary information underused. Therefore, we propose an innovative POI method that integrates attention-aware meta-paths based on deep (DNMF-AM)....

Understanding the balance between supply and demand of leisure services (LSs) in urban areas can benefit spatial planning improve quality life residents. In cities developing countries, pursuit rapid economic growth has ignored residents’ for LSs, thereby leading to a high short these services. However, due lack relevant research data, few studies have focused on mismatch LSs areas. As typical representatives multisource geographic social sensing data are readily available at various...

Point-of-interest (POI) recommendation is the prevalent personalized service in location-based social networks (LBSNs). A single use of matrix factorization (MF) or deep neural cannot effectively capture complex structure user–POI interactions. In addition, to alleviate data-sparsity problem, current methods primarily introduce auxiliary information users and POIs. Auxiliary often judged be equally valued, which will dissipate some valuable information. Hence, we propose a novel POI method...