Landslides are one of the most destructive natural hazards; they can drastically alter landscape morphology, destroy man-made structures, and endanger people's life. Landslide susceptibility maps (LSMs), which show spatial likelihood landslide occurrence, crucial for environmental management, urban planning, minimizing economic losses. To date, majority research into data mining LSM uses small-scale case studies focusing on a single type landslide. This paper presents approach to producing...

Abstract Short‐term traffic flow prediction on a large‐scale road network is challenging due to the complex spatial–temporal dependencies, directed topology, and high computational cost. To address challenges, this article develops graph deep learning framework predict with accuracy efficiency. Specifically, we model dynamics of as an irreducible aperiodic Markov chain graph. Based representation, novel inception residual (STGI‐ResNet) developed for network‐based prediction. This integrates...

Short-term traffic forecasting on large street networks is significant in transportation and urban management, such as real-time route guidance congestion alleviation. Nevertheless, it very challenging to obtain high prediction accuracy with reasonable computational cost due the complex spatial dependency network time-varying patterns. To address these issues, this paper develops a residual graph convolution long short-term memory (RGC-LSTM) model for spatial-temporal data considering...

The study of urban spatial structure is currently one the most popular research fields in geography. This uses Lanzhou, major cities Northwest China, as a case area. Using industry classification POI data, nearest-neighbor index, kernel density estimation, and location entropy are adopted to analyze clustering-discrete distribution characteristics overall economic geographical elements city center, various elements, city. All these can provide scientific reference for sustainable...

The spatio-temporal residual network (ST-ResNet) leverages the power of deep learning (DL) for predicting volume citywide flows. However, this model, neglects dynamic dependency input flows in temporal dimension, which affects what features may be captured result. This study introduces a long short-term memory (LSTM) neural into ST-ResNet to form hybrid integrated-DL model predict volumes (called HIDLST). new can dynamically learn among via feedback connection LSTM improve accurate captures...

The content-based remote sensing image retrieval (CBRSIR) has recently become a hot topic due to its wide applications in analysis of data. However, since conventional CBRSIR is unsuitable harsh environments, this article focuses on the cross-modality (CM-CBRSIR) between synthetic aperture radar (SAR) and optical images. Besides large interclass small intraclass CBRSIR, CM-CBRSIR limited by prominent modality discrepancy caused different imaging mechanisms. To address limitation, study...

Prognostics is necessary to ensure the reliability and safety of lithium-ion batteries for hybrid electric vehicles or satellites. This process can be achieved by capacity estimation, which a direct fading indicator assessing state health battery. However, battery onboard difficult monitor. paper presents data-driven approach online estimation. First, six novel features are extracted from cyclic charge/discharge cycles used as indirect indicators. An adaptive multi-kernel relevance machine...

Recently, researchers have introduced deep learning methods such as convolutional neural networks (CNN) to model spatio-temporal data and achieved better results than those with conventional methods. However, these CNN-based models employ a grid map represent spatial data, which is unsuitable for road-network-based data. To address this problem, we propose residual network modeling (DSTR-RNet). The proposed constructs locally-connected layers (LCNR) road topology integrates the dependency....

Existing deepfake analysis methods are primarily based on discriminative models, which significantly limit their application scenarios. This paper aims to explore interactive by performing instruction tuning multi-modal large language models (MLLMs). will face challenges such as the lack of datasets and benchmarks, low training efficiency. To address these issues, we introduce (1) a GPT-assisted data construction process resulting in an instruction-following dataset called DFA-Instruct, (2)...

Unlike other countries, sole traders in China have conventionally been defined law terms of “households” or hu, which there are two types: individual industrial and commercial households rural contractual management households. The former currently the main legal towns cities, while latter mainly farmers areas who engage business. concept household, as vehicle for types trader China, stresses collective, group, family nature is not simple sum interests. This article explores emergence...

The rapid advancements in Autonomous Driving Systems (ADS) have necessitated robust software testing to ensure safety and reliability. However, automating the generation of scalable concrete test scenarios remains a significant challenge. Current scenario-based case methods often face limitations, such as unrealistic scenes inaccurate vehicle trajectories. These challenges largely result from loss map information during data extraction lack an effective verification mechanism mitigate...

With the ubiquity of GPS-enabled devices and location-based social network services, research on human mobility becomes quantitatively achievable. Understanding it could lead to appealing applications such as city planning epidemiology. In this paper, we focus predicting whether two individuals are friends based their information. Intuitively, tend visit similar places, thus number co-occurrences should be a strong indicator friendship. Besides, visiting time interval between users also has...

Understanding the employment status of passengers in public transit systems is significant for transport operators many real applications such as forecasting travel demand and providing personalized transportation service. This paper develops a deep learning approach to infer passenger's by using smart card data (SCD) with household survey. first extracts an individual weekly patterns different modes from raw SCD three-dimensional image. A architecture, called thresholding multi-channel...

Predicting trip purpose from comprehensive and continuous smart card data is beneficial for transport city planners in investigating travel behaviors urban mobility. Here, we propose a framework, ActivityNET, using Machine Learning (ML) algorithms to predict passengers' Smart Card (SC) Points-of-Interest (POIs) data. The feasibility of the framework demonstrated two phases. Phase I focuses on extracting activities individuals' daily patterns combining them with POIs proposed "activity-POIs...

The large volume of data automatically collected by smart card fare systems offers a rich source information regarding daily human activities with high resolution spatial and temporal representation. This provides an opportunity for aiding transport planners policy-makers to plan cities more responsively. However, there are currently limitations when it comes understanding the secondary individual commuters. Accordingly, in this paper, we propose framework detect infer from individuals'...

In the new era, vitality of urban space is an important engine development, and improvement core component spatial structure optimization renewal. However, availability data issue in evaluation vitality, continuous monitoring entire city difficult to achieve through traditional methods field research questionnaire interviews. Due this challenge, assessment have serious limitations analysis causes guidance development. Using mobile phone signaling data, study takes Changsha City as example...