Spatial tensors have been extensively used in a wide range of applications, including remote sensing, geospatial information systems, conservation planning, and urban planning. We study the problem of Spatially Compac...
详细信息
ISBN:
(纸本)9798400712456
Spatial tensors have been extensively used in a wide range of applications, including remote sensing, geospatial information systems, conservation planning, and urban planning. We study the problem of Spatially Compact Dense (SCD) block mining in a spatial tensor, which targets for discovering dense blocks that cover small spatial regions. However, most of existing dense block mining (DBM) algorithms cannot solve the SCD-block mining problem since they only focus on maximizing the density of candidate blocks, so that the discovered blocks are spatially loose, i.e., covering large spatial regions. Therefore, we first formulate the problem of mining top-k Spatially Compact Dense blocks (SCD-blocks) in spatial tensors, which ranks SCD-blocks based on a new scoring function that takes both the density value and the spatial coverage into account. Then, we adopt a filter-refinement framework that first generates candidate SCD-blocks with good scores in the filtering phase and then uses the traditional DBM algorithm to further maximize the density values of the candidates in the refinement phase. Due to the NP-hardness of the problem, we develop two types of solutions in the filtering phase, namely the top-down solution and the bottom-up solution, which can find good candidate SCD-blocks by approximately solving the new scoring function. The evaluations on four real datasets verify that compared with the dense blocks returned by existing DBM algorithms, the proposed solutions are able to find SCD-blocks with comparable density values and significantly smaller spatial coverage.
Fire detection algorithms, particularly those based on computer vision, encounter significant challenges such as high computational costs and delayed response times, which hinder their application in real-time systems...
详细信息
In multi-view multi-label learning (MVML), each sample can be represented by multiple view features and associated with multiple labels. Most existing MVML algorithms are based on the assumption that all views share t...
详细信息
ISBN:
(纸本)9798400712203
In multi-view multi-label learning (MVML), each sample can be represented by multiple view features and associated with multiple labels. Most existing MVML algorithms are based on the assumption that all views share the same label set. However, in practice, different views may contain distinct label information, that means a single view cannot fully represent all labels. Based on this issue, the LVSL algorithm effectively learns view-specific labels and obtains superior classification performance. However, LVSL still has the limitation that it fails to consider the correlations between views, leading to suboptimal learning results. In this paper, we propose an improved LVSL algorithm named LVSL_VC (LVSL with view consensus). We incorporate view consensus learning into the original LVSL framework. Firstly, we employ view weights to model view consensus, assuming that views with similar weights will yield similar prediction outputs, conversely, they will be different. Secondly, we integrate the view consensus into the LVSL framework and construct a new classification model. Finally, we utilize an alternating optimization method to solve the problem. Extensive experimental results demonstrate that the LVSL_VC outperforms other state-of-the-art MVML algorithms.
Refactoring is the process of restructuring existing code without changing its external behavior while improving its internal structure. Refactoring engines are integral components of modern Integrated Development Env...
详细信息
Analyzing sequential data is crucial in many domains, particularly due to the abundance of data collected from the Internet of Things paradigm. Time series classification, the task of categorizing sequential data, has...
详细信息
Frequent road incidents cause significant physical harm and economic losses globally. The key to ensuring road safety lies in accurately perceiving surrounding road incidents. However, the highly dynamic nature o...
详细信息
In recent years, the country has released a large number of standard documents related to prefabricated concrete components. Due to the dispersion and complexity of these standards, it is difficult for industry manage...
详细信息
Community detection in multiplex networks has emerged as a crucial research area due to its ability to capture complex interactions across multiple layers of interconnected data. Despite significant advancements, exis...
详细信息
Community detection in multiplex networks has emerged as a crucial research area due to its ability to capture complex interactions across multiple layers of interconnected data. Despite significant advancements, existing methods often face critical challenges, including computational time, resolution limit, free parameter tuning, training models, etc. To overcome these limitations, this paper presents LCDMN (Layer-Coupled Diffusion for Multiplex Networks) algorithm designed for accurate and efficient community detection in multiplex networks. LCDMN employs dynamic scaling and layer coupling to adaptively identify community structures across diverse network configurations, offering improved resilience to network's density and structural ambiguity. LCDMN addresses the challenges of layer diversity by: (1) dynamically weighting layers based on critical parameters such as layer correlation, layer nodes activity variance, and attractiveness, (2) developing a robust node scoring method, (3) the aggregating layers of multiplex network into a single-layer, weighted graph, (4) employing a label diffusion approach with mechanisms for handling overlapping nodes, and (5) refining community structures through a dynamic merging process that adaptively adjusts layer contributions and community boundaries during execution, ensuring context-sensitive resolution of structural ambiguity. Nodes and edges are scored using network topology and structural metrics to efficiently incorporate in label diffusion process for detecting initial communities. The approach balances computational efficiency with precision, enabling the detection of cohesive and well-defined communities in complex networks. Experimental evaluations on real-world and synthetic multiplex networks demonstrate that LCDMN consistently outperforms state-of-the-art methods, such as Infomap, MDLPA, MPBTV, LART, DGFM3 and GenLouvain, in terms of Normalized Mutual Information (NMI), Adjusted Rand Index (ARI), and modularity.
Language-based text provide valuable insights into people’s lived experiences. While traditional qualitative analysis is used to capture these nuances, new paradigms are needed to scale qualitative research effective...
详细信息
There is a growing concern about adversarial attacks against automatic speech recognition (ASR) systems. Although research into targeted universal adversarial examples (AEs) has progressed, current methods are constra...
详细信息
暂无评论