检索结果-内蒙古大学图书馆

Multiscale and Auto-Tuned Semi-Supervised Deep Subspace Clustering and Its Application in Brain Tumor Clustering

学校读者我要写书评

暂无评论

computers, Materials & Continua 2024年第6期79卷 4741-4762页

作者： Zhenyu Qian Yizhang Jiang Zhou Hong Lijun Huang Fengda Li Khin Wee Lai Kaijian Xia School of Artificial Intelligence and Computer Science Jiangnan UniversityWuxi214122China Department of Medical Imaging The Changshu Affiliated Hospital of Soochow UniversitySuzhou215500China Department of Neurosurgery The Changshu Affiliated Hospital of Soochow UniversityChangshu215500China Department of Scientific Research The Changshu Affiliated Hospital of Soochow UniversitySuzhou215500China Changshu Key Laboratory of Medical Artificial Intelligence and Big Data Suzhou215500China Department of Biomedical Engineering Faculty of EngineeringUniversity of MalayaKuala Lumpur50603Malaysia

In this paper,we introduce a novel Multi-scale and Auto-tuned Semi-supervised Deep Subspace Clustering(MAS-DSC)algorithm,aimed at addressing the challenges of deep subspace clustering in high-dimensional real-world data,particularly in the field of medical *** deep subspace clustering algorithms,which are mostly unsupervised,are limited in their ability to effectively utilize the inherent prior knowledge in medical *** MAS-DSC algorithm incorporates a semi-supervised learning framework that uses a small amount of labeled data to guide the clustering process,thereby enhancing the discriminative power of the feature ***,the multi-scale feature extraction mechanism is designed to adapt to the complexity of medical imaging data,resulting in more accurate clustering *** address the difficulty of hyperparameter selection in deep subspace clustering,this paper employs a Bayesian optimization algorithm for adaptive tuning of hyperparameters related to subspace clustering,prior knowledge constraints,and model loss *** experiments on standard clustering datasets,including ORL,Coil20,and Coil100,validate the effectiveness of the MAS-DSC *** results show that with its multi-scale network structure and Bayesian hyperparameter optimization,MAS-DSC achieves excellent clustering results on these ***,tests on a brain tumor dataset demonstrate the robustness of the algorithm and its ability to leverage prior knowledge for efficient feature extraction and enhanced clustering performance within a semi-supervised learning framework.

关键词： Deep subspace clustering multiscale network structure automatic hyperparameter tuning semi-supervised medical image clustering

An Experimental Evaluation of Imputation Models for Spatial-Temporal Traffic data

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Guo, Shengnan Wei, Tonglong Huang, Yiheng Zhao, Miaomiao Chen, Ran Lin, Yan Lin, Youfang Wan, Huaiyu School of Computer Science and Technology Beijing Jiaotong University Beijing China Beijing Key Laboratory of Traffic Data Analysis and Mining Beijing China Department of Computer Science Aalborg University Aalborg Denmark

Traffic data imputation is a critical preprocessing step in intelligent transportation systems, enabling advanced transportation services. Despite significant advancements in this field, selecting the most suitable model for practical applications remains challenging due to three key issues: 1) incomprehensive consideration of missing patterns that describe how data loss along spatial and temporal dimensions, 2) the lack of test on standardized datasets, and 3) insufficient evaluations. To this end, we first propose practice-oriented taxonomies for missing patterns and imputation models, systematically identifying all possible forms of real-world traffic data loss and analyzing the characteristics of existing models. Furthermore, we introduce a unified benchmarking pipeline to comprehensively evaluate 10 representative models across various missing patterns and rates. This work aims to provide a holistic understanding of traffic data imputation research and serve as a practical guideline. Copyright © 2024, The Authors. All rights reserved.

关键词： Taxonomies

Multiplex measures for higher-order networks

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Lotito, Quintino Francesco Montresor, Alberto Battiston, Federico Department of Information Engineering and Computer Science University of Trento via Sommarive 9 Trento38123 Italy Department of Network and Data Science Central European University Vienna1100 Austria

A wide variety of complex systems are characterized by interactions of different types involving varying numbers of units. Multiplex hypergraphs serve as a tool to describe such structures, capturing distinct types of higher-order interactions among a collection of units. In this work, we introduce a comprehensive set of measures to describe structural connectivity patterns in multiplex hypergraphs, considering scales from node and hyperedge levels to the system’s mesoscale. We validate our measures with three real-world datasets: scientific co-authorship in physics, movie collaborations, and high school interactions. This validation reveals new collaboration patterns, identifies trends within and across movie subfields, and provides insights into daily interaction dynamics. Our framework aims to offer a more nuanced characterization of real-world systems marked by both multiplex and higher-order interactions. Copyright © 2024, The Authors. All rights reserved.

关键词： Large scale systems

Provably Optimal Memory Capacity for Modern Hopfield Models: Transformer-Compatible Dense Associative Memories as Spherical Codes

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Hu, Jerry Yao-Chieh Wu, Dennis Liu, Han Center for Foundation Models and Generative AI United States Department of Computer Science United States Department of Statistics and Data Science Northwestern University EvanstonIL60208 United States

We study the optimal memorization capacity of modern Hopfield models and Kernelized Hopfield Models (KHMs), a transformer-compatible class of Dense Associative Memories. We present a tight analysis by establishing a connection between the memory configuration of KHMs and spherical codes from information theory. Specifically, we treat the stored memory set as a specialized spherical code. This enables us to cast the memorization problem in KHMs into a point arrangement problem on a hypersphere. We show that the optimal capacity of KHMs occurs when the feature space allows memories to form an optimal spherical code. This unique perspective leads to: (i) An analysis of how KHMs achieve optimal memory capacity, and identify corresponding necessary conditions. Importantly, we establish an upper capacity bound that matches the well-known exponential lower bound in the literature. This provides the first tight and optimal asymptotic memory capacity for modern Hopfield models. (ii) A sub-linear time algorithm U-Hop+ to reach KHMs’ optimal capacity. (iii) An analysis of the scaling behavior of the required feature dimension relative to the number of stored memories. These efforts improve both the retrieval capability of KHMs and the representation learning of corresponding transformers. Experimentally, we provide thorough numerical results to back up theoretical findings. © 2024, CC BY-NC-SA.

关键词： Associative storage

Regularized Adaptive Momentum Dual Averaging with an Efficient Inexact Subproblem Solver for Training Structured Neural Network

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Huang, Zih-Syuan Lee, Ching-Pei Department of Computer Science and Information Engineering National Taiwan University Taipei106 Taiwan Department of Advanced Data Science Institute of Statistical Mathematics Tachikawa Tokyo190-8562 Japan

We propose a Regularized Adaptive Momentum Dual Averaging (RAMDA) algorithm for training structured neural networks. Similar to existing regularized adaptive methods, the subproblem for computing the update direction of RAMDA involves a nonsmooth regularizer and a diagonal preconditioner, and therefore does not possess a closed-form solution in general. We thus also carefully devise an implementable inexactness condition that retains convergence guarantees similar to the exact versions, and propose a companion efficient solver for the subproblems of both RAMDA and existing methods to make them practically feasible. We leverage the theory of manifold identification in variational analysis to show that, even in the presence of such inexactness, the iterates of RAMDA attain the ideal structure induced by the regularizer at the stationary point of asymptotic convergence. This structure is locally optimal near the point of convergence, so RAMDA is guaranteed to obtain the best structure possible among all methods converging to the same point, making it the first regularized adaptive method outputting models that possess outstanding predictive performance while being (locally) optimally structured. Extensive numerical experiments in large-scale modern computer vision, language modeling, and speech tasks show that the proposed RAMDA is efficient and consistently outperforms state of the art for training structured neural network. Implementation of our algorithm is available at https://***/ismoptgroup/RAMDA/. © 2024, CC BY.

关键词： HTTP

Integrated VGG19 and Gated Recurrent Unit with Cosine Similarity Metric for Content based Image Retrieval

学校读者我要写书评

暂无评论

Integrated VGG19 and Gated Recurrent Unit with Cosine Simila...

Software, Systems and Information Technology (SSITCON), International Conference on

作者： Rajath Arakere Narayanaswamy Vidyalakshmi Krishne Gowda Department of Computer Science & Engineering GSSS Institute of Engineering & Technology for Women Mysuru India Sri Siddhartha Institute of Technology SSAHE University Tumakuru India Department of Computer Science & Engineering (Data Science) Sri Siddhartha Institute of Technology SSAHE University Tumakuru India

ISBN: (数字)9798350352931

ISBN: (纸本)9798350352948

Technological advancement has provided huge databases for different genres of image to obtain efficient visual information to satisfy users. The existing algorithms tried to extract the important feature vector from the images to analyze the similarities between the query image and dataset image but failed to reach their expected results. The noises generated during model training, overfitting and gradient vanishing problems are the problems faced by the existing models. To overcome the problems of the existing methods, the VGG19 and Gated Recurrent Unit (GRU) with cosine similarity are proposed for content-based image retrieval. The z-score normalization technique was used to standardize the intensity of the images which enhanced the training of the model. The VGG19 algorithm extracted spatial features and the GRU model extracted the temporal features from both database and query images. The cosine similarity metric was used for calculating the similarities of both the images and retrieved based on the feature vector similarities. The proposed VGG19 and GRU with cosine similarity metric has obtained a better average precision of 98.89% for Corel 1 K dataset and 84.90% for Corel 10 K dataset compared to the existing Hybrid Graph-based Gray Level Co-Occurrence Matrix (HGGLCM) model.

关键词： Measurement Training Visualization Image retrieval Logic gates Feature extraction Vectors Spatial databases Visual databases Overfitting

Efficient and Accurate Learning of Mixtures of Plackett-Luce Models

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Nguyen, Duc Zhang, Anderson Y. Depart of Computer and Information Science University of Pennsylvania United States Department of Statistics and Data Science University of Pennsylvania United States

Mixture models of Plackett-Luce (PL) - one of the most fundamental ranking models - are an active research area of both theoretical and practical significance. Most previously proposed parameter estimation algorithms instantiate the EM algorithm, often with random initialization. However, such an initialization scheme may not yield a good initial estimate and the algorithms require multiple restarts, incurring a large time complexity. As for the EM procedure, while the E-step can be performed efficiently, maximizing the log-likelihood in the M-step is difficult due to the combinatorial nature of the PL likelihood function (Gormley and Murphy 2008). Therefore, previous authors favor algorithms that maximize surrogate likelihood functions (Zhao et al. 2018, 2020). However, the final estimate may deviate from the true maximum likelihood estimate as a consequence. In this paper, we address these known limitations. We propose an initialization algorithm that can provide a provably accurate initial estimate and an EM algorithm that maximizes the true log-likelihood function efficiently. Experiments on both synthetic and real datasets show that our algorithm is competitive in terms of accuracy and speed to baseline algorithms, especially on datasets with a large number of items. © 2023, CC BY.

关键词： Large dataset

Feasibility Study of Semi-supervised Intrusion Detection Systems for Industrial Control Systems 10

学校读者我要写书评

暂无评论

Feasibility Study of Semi-supervised Intrusion Detection Sys...

10th International Conference on Applied System Innovation, ICASI 2024

作者： Loo, Fu-Nie Chen, Tung-Shou Tsai, Chia-Wei Lin, Jason Yang, Chun-Wei National Taichung University of Science and Technology Department of Computer Science and Information Engineering No.129 Sec.3 Sanmin Rd. North Dist. Taichung City404 Taiwan National Chung Hsing University Department of Computer Science and Engineering 145 Xingda Rd. South Dist. Taichung City402202 Taiwan China Medical University Master Program for Digital Health Innovation No. 100 Sec. 1 Jingmao Rd. Beitun Dist. Taichung City406040 Taiwan

ISBN: (纸本)9798350394924

Industrial Control systems (ICS) automate industrial processes but also introduces cybersecurity threats. Intrusion Detection System (IDS) are crucial for detecting cyber-attacks on ICS, yet zero-day attacks are often inefficiently detecting with supervised learning. This study employs semi-supervised learning using one-class SVM, isolation forest, and Local Outlier Factor (LOF), to train IDS models. Utilizing dataset collected from a self-build virtual ICS environment, the study demonstrates the feasibility of these models in detecting common attack like Injection, ARP, and Man-in-the-Middle. © 2024 IEEE.

关键词： Zero-day attack

Augmented Reality Immersive World with Hologram Special Effect in Early Childhood Education 2nd

学校读者我要写书评

暂无评论

Augmented Reality Immersive World with Hologram Special Effe...

2nd International Conference on science, Engineering Management and Information Technology, SEMIT 2023

作者： Astriani, Maria Seraphina Bahana, Raymond Ahmad, Arif Priyono Susilo Kurniawan, Andreas Yi, Lee Huey Computer Science Department School of Computing and Creative Arts Bina Nusantara University Jakarta11480 Indonesia Creative Advertising Program Visual Communication Design Department School of Design Bina Nusantara University Jakarta11480 Indonesia Water Lily Melaka75200 Malaysia

ISBN: (纸本)9783031722837

Early childhood education is a critical stage of education because it may establish the foundation for an individual’s intelligence. Students of this age group are generally unable to remain seated and study for long periods of time because they get bored easily. They are more likely to be attracted to visual information and enjoy interactive activities. This condition challenges the world of education to keep developing learning methods that are able to help students to learn the materials. The combination of Augmented Reality technology and hologram special effects is a promising solution because it allows students to see an immersive world and allows them to interact with real-life physical objects while learning. Augmented Reality immersive world solution with hologram special effects in early childhood education has successfully been tested and implemented by using two smartphones, one acts as an input-camera and the other acts as an output to display the hologram. This solution is specially designed to make it easier for students to learn the materials without using wearable devices. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

关键词： Holograms