Abstract.With the increasing traffic demand, the trade-off between freeway traffic flow and driving safety has become increasingly prominent. Traditional right-driving rules often fail to balance driving safety and road network efficiency under varying traffic densities. This paper aims to deeply explore the impact mechanism of freeway driving rules on the spatio-temporal evolution of traffic flow. First, by introducing the concept of safe following distance, a vehicle anti-collision model under different driving states and overtaking behaviors is constructed to quantify the safety risks in microscopic … Read more →

Abstract.Based on data from the Global Innovation Index, this paper constructs an integrated data-driven framework that combines feature selection, cluster analysis, and predictive modeling. First, a random forest model is used to assess the importance of multidimensional indicators, identifying key variables from high-dimensional features to effectively capture complex nonlinear relationships. Second, the K-means method is employed to stratify the innovation levels of different countries, achieving structured classification. Building on this foundation, a sliding-window-based random forest regression … Read more →

Abstract.Amyotrophic lateral sclerosis (ALS) has evolved in clinical understanding from an isolated motor neuron disease into a complex multisystem disorder. Current evidence firmly establishes cognitive deficits and behavioral dysfunctions as inherent, non-motor elements of ALS, factors that drastically alter clinical outcomes, life expectancy, and caregiver stress. In light of this growing recognition, multimodal magnetic resonance imaging (MRI) has emerged as a crucial tool for elucidating the multiscale neuropathology of these symptoms, capturing a trajectory from structural damage to functional … Read more →

Abstract.The digital economy's unprecedented expansion has been accompanied by increasingly sophisticated, networked fraudulent activities that exploit systemic vulnerabilities. Traditional tabular machine learning models evaluate entities in isolation, often failing to detect the collusive, multi-step strategies employed by modern adversaries. Consequently, Graph Neural Networks (GNNs) have emerged as the foundational architecture for advanced fraud detection, leveraging complex topological relationships to uncover illicit patterns. This paper systematically traces the evolutionary trajectory from … Read more →

Abstract.Haplotype assembly aims to partition sequencing reads according to their chromosomal origin based on variant information and to reconstruct the allelic configurations along individual chromosomes. Its core significance lies in revealing the true combinations of variants on the same chromosome, thereby enhancing the biological interpretability of genetic information. This paper presents a systematic review of haplotype assembly methods, with a focus on their underlying models and algorithmic strategies. Existing approaches are broadly categorized into reference-guided and de novo methods, and … Read more →

Abstract.Medical image analysis plays a critical role in modern clinical diagnosis and treatment planning. While deep learning, particularly convolutional neural networks, has achieved remarkable success, conventional models struggle to capture the complex high-order relationships inherent in anatomical structures and pathological patterns. Hypergraph neural networks (HGNNs), which generalize standard graphs by allowing hyperedges to connect multiple nodes, have emerged as a powerful paradigm for modeling such high-order interactions. This review provides a systematic overview of hypergraph learning … Read more →

Abstract.This study investigated how differences in consumer perceptions of content origin—specifically between Gen-AI and human creation—influence brand culture, with a focus on addressing the underexplored connection between content authorship and brand cultural formation. Drawing on the Elaboration Likelihood Model (ELM), brand signal theory, and cultural consistency theory, two empirical studies were carried out to unpack the cognitive and affective mechanisms underlying this relationship. Study 1 compared consumer perceptions across four dimensions—brand authenticity, credibility, electronic … Read more →

Abstract.In recent years, the application of large language models (LLMs) in intelligent cockpits has been continuously deepened, significantly improving the natural human-vehicle interaction experience. However, the generative characteristics of models, multi-modal data interaction, and access to vehicle control privileges have also introduced a series of new information security vulnerabilities such as prompt injection, data privacy leakage, unauthorized vehicle manipulation, and model hallucinations. Traditional in-vehicle security testing systems mostly target vulnerabilities in in-vehicle … Read more →

Abstract.As a significant breakthrough in recent years for artificial intelligence research, multimodal large-scale language models have extended the ability of language models across multiple dimensions, including vision and hearing, into non-linguistic forms through technical processes such as cross-modal alignment. Examine the mechanisms, problems and future paths for cross-modal alignment in multimodal large-scale language models based on representations learned by using the theory of representation learning, transformer-based attention mechanism and information-theoretical framework as core … Read more →

Abstract.Coalbed methane (CBM) content is a critical indicator for CBM resource evaluation and development. Traditional prediction methods rely mainly on core experiments or empirical formulas, which suffer from high costs, long cycles, limited generalization ability, and difficulties in capturing the complex nonlinear relationships between geological/geophysical parameters and CBM content. To address these shortcomings, this paper proposes an intelligent prediction method for deep CBM content based on random forest regression and professional geological feature engineering. The method integrates … Read more →

Abstract.Multi-step time series forecasting based on iterative strategies often suffers from severe error accumulation, which significantly degrades long-horizon prediction performance. To address this issue, this paper proposes a hybrid forecasting framework that integrates granular computing with Long Short-Term Memory (LSTM) networks. First, the original time series is transformed into a sequence of temporal information granules, each represented by a triplet consisting of cardinality, mean, and standard deviation, which respectively characterize data density, trend level, and fluctuation intensity. … Read more →

Abstract.Accurate retrieval of farmland soil moisture from multi-source remote sensing data requires simultaneous consideration of time-series interpolation strategies and machine learning algorithm selection. However, the theoretical basis explaining how environmental conditions modulate their joint effects remains insufficiently articulated. This paper proposes a unified conceptual framework termed the "Interpolation-Algorithm-Environment" (IAE) triad, which explains why no universally optimal retrieval pipeline exists and why optimal choices are inherently environment-contingent. We further … Read more →

Abstract.This paper presents a multimodal conversational teaching agent that co-creates classroom slide decks and lesson plans with teachers through a staged review process. The prototype was motivated by a practical teaching-design scenario in which courseware preparation is fragmented across several tools and teacher intent is often captured only superficially. The proposed system combines three capabilities: multi-turn intent clarification, retrieval-augmented generation over a locally curated knowledge base, and optional reference-file grounding from uploaded PDF, DOCX, PPT, image, or video … Read more →

Abstract.Retail product packaging text detection is crucial for intelligent label auditing and smart retail applications. While deep learning models, particularly the YOLO series, have demonstrated remarkable real-time inference capabilities, detecting text in complex retail scenarios presents formidable challenges due to extreme scale variations, dense distributions, and severe background interference. During the repeated downsampling processes in standard YOLO architectures, the fine-grained features of microscopic text are prone to degradation, and introducing conventional global attention … Read more →

Abstract.The eukaryotic genome exists in a highly folded three-dimensional structure within the nucleus. As the core functional unit of the 3D genome, chromatin loops are formed via the loop extrusion mechanism mediated by CTCF proteins and cohesin complexes. They directly facilitate the spatial interaction between distal enhancers and gene promoters, and precisely regulate essential biological processes including gene transcription, DNA replication and damage repair. Structural abnormalities of chromatin loops are closely associated with the occurrence and progression of numerous diseases, such as … Read more →

Abstract.Accurate classification of chromatin A/B compartments is one of the core issues in three-dimensional genomics, which is critical for understanding how genome spatial organization affects gene regulation, cell fate determination, and disease pathogenesis. Since Hi C technology first revealed in 2009 that the genome is spatially segregated into transcriptionally active A compartments and inactive B compartments, principal component analysis (PCA) based on Hi C contact matrices has long served as the standard strategy for compartment identification. However, with the in-depth advancement of … Read more →