检索结果-内蒙古大学图书馆

International Conference on machine learning and Applications (ICMLA)

作者： Simon Klüttermann Jérôme Rutinowski Frederik Polachowski Anh Nguyen Britta Grimme Moritz Roidl Emmanuel Müller TU Dortmund University Dortmund Germany Lamarr Institute for Machine Learning and Artificial Intelligence Dortmund Germany Paderborn University Paderborn Germany Research Center Trustworthy Data Science and Security Dortmund Germany

ISBN: (数字)9798350374889

ISBN: (纸本)9798350374896

In this contribution, we introduce a novel ensemble method for the re-identification of industrial entities, using images of chipwood pallets and galvanized metal plates as dataset examples. Our algorithms replace commonly used, complex siamese neural networks with an ensemble of simplified, rudimentary models, providing wider applicability, especially in hardware-restricted scenarios. Each ensemble sub-model uses different types of extracted features of the given data as its input, allowing for the creation of effective ensembles in a fraction of the training duration needed for more complex state-of-the-art models. We reach state-of-the-art performance at our task, with a Rank-1 accuracy of over 77% and a Rank-10 accuracy of over 99%, and introduce five distinct feature extraction approaches, and study their combination using different ensemble methods.

关键词： Training Accuracy Neural networks Metals Feature extraction Silicon Ensemble learning Security Reliability Pallets

来源：评论

学校读者我要写书评

暂无评论

Turnstile ℓp leverage score sampling with applications 24

Turnstile ℓp leverage score sampling with applications

引用

Proceedings of the 41st International Conference on machine learning

作者： Alexander Munteanu Simon Omlor Dortmund Data Science Center Faculties of Statistics and Computer Science TU Dortmund University Dortmund Germany Faculty of Statistics TU Dortmund University Dortmund Germany and Lamarr-Institute for Machine Learning and Artificial Intelligence Dortmund Germany

The turnstile data stream model offers the most flexible framework where data can be manipulated dynamically, i.e., rows, columns, and even single entries of an input matrix can be added, deleted, or updated multiple times in a data stream. We develop a novel algorithm for sampling rows ai of a matrix A ∈ ℝn×d, proportional to their ℓp norm, when A is presented in a turnstile data stream. Our algorithm not only returns the set of sampled row indexes, it also returns slightly perturbed rows ãi ≈ ai, and approximates their sampling probabilities up to e relative error. When combined with preconditioning techniques, our algorithm extends to ℓp leverage score sampling over turnstile data streams. With these properties in place, it allows us to simulate subsampling constructions of coresets for important regression problems to operate over turnstile data streams with very little overhead compared to their respective off-line subsampling algorithms. For logistic regression, our framework yields the first algorithm that achieves a (1 + ε) approximation and works in a turnstile data stream using polynomial sketch/subsample size, improving over O(1) approximations, or exp(1/ε) sketch size of previous work. We compare experimentally to plain oblivious sketching and plain leverage score sampling algorithms for ℓp and logistic regression.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Topological Analysis on Multi-scenario Graphs: Applications Toward Discerning Variability in SARS-CoV-2 and Topic Similarity in Research

引用

Transactions of the Indian National Academy of Engineering : an international journal of engineering and technology 2022年第1期7卷 365-374页

作者： Sourav Biswas Malay Bhattacharyya Sanghamitra Bandyopadhyay Indian Statistical Institute Kolkata India. University of Calcutta Kolkata India. Machine Intelligence Unit Centre for Artificial Intelligence and Machine Learning Technology Innovation Hub on Data Science Big Data Analytics and Data Curation Indian Statistical Institute Kolkata India.

A network is often an obvious choice for modeling real-life interconnected systems, where the nodes represent interacting objects and the edges represent their associations. There has been immense progress in complex network analysis with methods and tools that can provide important insights into the respective scenario. In the advancement of information technology and globalization, the amount of data is increasing day by day, and it is indeed incomprehensible without the help of network science. This work highlights how we can model multiple interaction scenarios under a single umbrella to uncover novel insights. We show that a varying scenario gets reflected by the change of topological patterns in interaction networks. We construct multi-scenario graphs, a novel framework proposed by us, from real-life environments followed by topological analysis. We focus on two different application areas: analyzing geographical variations in SARS-CoV-2 and studying topic similarity in citation patterns.

关键词： Citation Network Graphlet and Graphlet Degree Distribution Multi-scenario Graphs SARS-CoV-2 Topic Similarity Network Topological Analysis

来源：评论

学校读者我要写书评

暂无评论

Sample Complexity of Posted Pricing for a Single Item

arXiv

引用

arXiv 2024年

作者： Jin, Billy Kesselheim, Thomas Ma, Will Singla, Sahil University of Chicago Booth School of Business United States University of Bonn Institute of Computer Science Lamarr Institute for Machine Learning and Artificial Intelligence Germany Columbia University Graduate School of Business Data Science Institute United States Georgia Tech School of Computer Science United States

Selling a single item to n self-interested buyers is a fundamental problem in economics, where the two objectives typically considered are welfare maximization and revenue maximization. Since the optimal mechanisms are often impractical and do not work for sequential buyers, posted pricing mechanisms, where fixed prices are set for the item for different buyers, have emerged as a practical and effective alternative. This paper investigates how many samples are needed from buyers’ value distributions to find near-optimal posted prices, considering both independent and correlated buyer distributions, and welfare versus revenue maximization. We obtain matching upper and lower bounds (up to logarithmic factors) on the sample complexity for all these settings. Copyright © 2024, The Authors. All rights reserved.

关键词： Sales

来源：评论

学校读者我要写书评

暂无评论

Position: insights from survey methodology can improve training data 24

Position: insights from survey methodology can improve train...

引用

Proceedings of the 41st International Conference on machine learning

作者： Stephanie Eckman Barbara Plank Frauke Kreuter Social Data Science Center University of Maryland College Park MD Center for Information and Language Processing (CIS) LMU Munich Germany and Computer Science Department IT University of Copenhagen Denmark and Munich Center for Machine Learning (MCML) LMU Munich Germany Institute for Statistics and Munich Center for Machine Learning (MCML) LMU Munich Germany and Social Data Science Center and Joint Program in Survey Methodology University of Maryland College Park MD

Whether future AI models are fair, trustworthy, and aligned with the public's interests rests in part on our ability to collect accurate data about what we want the models to do. However, collecting high-quality data is difficult, and few AI/ML researchers are trained in data collection methods. Recent research in data-centric AI has show that higher quality training data leads to better performing models, making this the right moment to introduce AI/ML researchers to the field of survey methodology, the science of data collection. We summarize insights from the survey methodology literature and discuss how they can improve the quality of training and feedback data. We also suggest collaborative research ideas into how biases in data collection can be mitigated, making models more accurate and human-centric.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Addressing data Imbalance in machine learning: Challenges and Approaches

Addressing Data Imbalance in Machine Learning: Challenges an...

引用

Mobile Computing and Sustainable Informatics (ICMCSI), International Conference on

作者： Meher Langote Nikita Zade Swapnil Gundewar Artificial Intelligence and Data Science Faculty of Engineering and Technology Datta Meghe Institute of Higher Education and Research Sawangi Maharashtra Artificial Intelligence and Machine Learning Faculty of Engineering and Technology Datta Meghe Institute of Higher Education and Research Sawangi Maharashtra

ISBN: (数字)9798331522667

ISBN: (纸本)9798331522674

A key difficulty of dealing with imbalanced datasets is that machine learning models are trained to favor the majority class and therefore perform poorly when it comes to a minority class classification. In this paper, we consider the following methods of tackling the issue of the overrepresentation of a class; random oversampling, under-sampling, SMOTE, cost-sensitive learning, and ensemble methods. Comparing them, we assess the performance of each of these techniques within different classifiers. Techniques including SMOTE together with combined methods including Random Forests also show higher capacity in minority class predictions. However, problems such as overfitting and the increase in the computational load are not escapable yet. We briefly recap the strengths and weaknesses of each approach and propose potential refinements including a combination of both methods and dynamic resampling methods to serve as the next steps toward improving the model. Thereby, our results draw attention to the necessity of accurate methods for handling the class imbalance, which will further yield more accurate and fair machine learning solutions across a variety of applications.

关键词： Adaptation models Accuracy machine learning algorithms Computational modeling Ensemble learning Informatics Random forests Mobile computing Overfitting Load modeling

来源：评论

学校读者我要写书评

暂无评论

AEON: Adaptive Estimation of Instance-Dependent In-Distribution and Out-of-Distribution Label Noise for Robust learning

arXiv

引用

arXiv 2025年

作者： Garg, Arpit Nguyen, Cuong Felix, Rafael Liu, Yuyuan Do, Thanh-Toan Carneiro, Gustavo Australian Institute for Machine Learning University of Adelaide Australia Centre for Vision Speech and Signal Processing University of Surrey United Kingdom Department of Engineering Science University of Oxford United Kingdom Department of Data Science and AI Monash University Australia

Robust training with noisy labels is a critical challenge in image classification, offering the potential to reduce reliance on costly clean-label datasets. Real-world datasets often contain a mix of in-distribution (ID) and out-of-distribution (OOD) instance-dependent label noise, a challenge that is rarely addressed simultaneously by existing methods and is further compounded by the lack of comprehensive benchmarking datasets. Furthermore, even though current noisy-label learning approaches attempt to find noisy-label samples during training, these methods do not aim to estimate ID and OOD noise rates to promote their effectiveness in the selection of such noisy-label samples, and they are often represented by inefficient multi-stage learning algorithms. We propose the Adaptive Estimation of Instance-Dependent In-Distribution and Out-of-Distribution Label Noise (AEON) approach to address these research gaps. AEON is an efficient one-stage noisy-label learning methodology that dynamically estimates instance-dependent ID and OOD label noise rates to enhance robustness to complex noise settings. Additionally, we introduce a new benchmark reflecting real-world ID and OOD noise scenarios. Experiments demonstrate that AEON achieves state-of-the-art performance on both synthetic and real-world datasets*. © 2025, CC BY-NC-ND.

关键词： Contrastive learning

来源：评论

学校读者我要写书评

暂无评论

machine learning and Deep learning Approaches for Automated Diabetic Retinopathy Diagnosis

Machine Learning and Deep Learning Approaches for Automated ...

引用

Intelligent Systems for Cybersecurity (ISCS), International Conference on

作者： Bhushan Fulkar Rushikesh Burle Pawan Patil Sanskruti Gaurkhade Swapnil Gundewar Utkarsha Pacharaney Department of Artificial Intelligence and Data science Datta Meghe Institute of Higher Education and Research Wardha India Department of Artificial Intelligence and Machine Learning Datta Meghe Institute of Higher Education and Research(DU) Wardha Maharashtra Department of Computer Science and Design Datta Meghe Institute of Higher Education and Research Wardha India Department of Artificial Intelligence and Machine Learning Datta Meghe Institute of Higher Education and Research Wardha India

ISBN: (数字)9798350375237

ISBN: (纸本)9798350375244

Since diabetic retinopathy (DR) is the primary cause of blindness in working-age adults worldwide, vision loss must be avoided by receiving early detection and treatment. In recent years, deep learning techniques have become more and more useful tools for automated drug discovery, possibly offering solutions to overcome the limitations of traditional screening methods. In-depth descriptions of the most recent deep learning methods for DR detection are provided in this review article. We begin by discussing the importance of DR screening, its prevalence, and the challenges faced by healthcare systems in managing this condition. Next, we discuss different deep learning architectures used for DR detection, such as CNNs and RNNs, and their combinations. We investigate how to improve the precision, resilience, and interpretability of DR detection models through the use of explainable AI, ensemble methods, transfer learning, and attention mechanisms. We also examine publicly available datasets, evaluation metrics, and preprocessing techniques that are commonly used to benchmark DR detection algorithms. We discuss the limitations and possible uses of deep learning-based deep readout detection, emphasizing the significance of large-scale, diverse datasets, the interpretability of the model, and real-world deployment concerns. The overall objectives of this review are to advance the field of automated DR detection by highlighting recent developments, compiling existing knowledge, and identifying areas that require further research.

关键词： Deep learning Diabetic retinopathy Reviews Transfer learning Imaging Transforms Medical services

来源：评论

学校读者我要写书评

暂无评论

MSI-UNet: A Flexible UNet-Based Multi-Scale Interactive Framework for 3D Gastric Tumor Segmentation on CT Scans

MSI-UNet: A Flexible UNet-Based Multi-Scale Interactive Fram...

引用

IEEE International Symposium on Biomedical Imaging

作者： Heyun Chen Zifan Chen Jie Zhao Haoshen Li Jiazheng Li Yiting Liu Mingze Yuan Peng Bao Xinyu Nan Bin Dong Lei Tang Li Zhang Center for Data Science Peking University China National Engineering Laboratory for Big Data Analysis and Applications Peking University China Peking University Changsha Institute for Computing and Digital Economy China Peking University Cancer Hospital&Institute China Beijing International Center for Mathematical Research Peking University China Center for Machine Learning Research Peking University China

ISBN: (数字)9798350313338

ISBN: (纸本)9798350313345

Accurate segmentation of gastric tumors is critical yet presents a formidable challenge in medical imaging, where conventional UNet-based frameworks, despite their prevalence, falter on intricate tumor samples due to their limited interactive capacities. The SAM-based segmentation methods address this shortcoming yet with insufficient accuracy. By ingeniously blending images with mask inputs, our MSI-UNet leverages a U-shaped design to deliver pixel-perfect segmentation accuracy, while a novel multi-scale attention module adeptly harnesses interaction points for refined information extraction. When benchmarked on gastric tumor segmentation tasks, MSI-UNet surpasses existing state-of-the-art methods, elevating the Dice Similarity Coefficient (DSC) from 74.82% to 79.3% and minimizing Average Surface Distance (ASD) from 6.46 to 1.98, achieving a comparable accuracy with inter-radiologist consistency of 79.7% DSC. Furthermore, our framework demonstrates superior predictive prowess in survival analysis, enhancing the C-index from 61.7% to 68.68%. Ample experimental comparisons have substantiated that MSI-UNet holds the potential to offer considerable assistance to healthcare professionals in managing and decoding subsequent medical procedures.

关键词： Image segmentation Accuracy Three-dimensional displays Medical services Predictive models Information retrieval Decoding

来源：评论

学校读者我要写书评

暂无评论

machine-learning Force Fields Reveal Shallow Electronic States on Dynamic Halide Perovskite Surfaces

arXiv

引用

arXiv 2025年

作者： Delgado, Frederico P. Simões, Frederico Kronik, Leeor Kaiser, Waldemar Egger, David A. Physics Department TUM School of Natural Sciences Technical University of Munich Germany Department of Molecular Chemistry and Materials Science Weizmann Institute of Science Israel Atomistic Modeling Center Munich Data Science Institute Technical University of Munich Germany Munich Center for Machine Learning Munich Germany

The spectacular performance of halide perovskites in optoelectronic devices is rooted in their favorable tolerance to structural defects. Previous studies showed that defects in these materials generate shallow electronic states that do not degrade device performance. However, how these shallow states persist amid the pronounced thermally-stimulated atomic dynamics on halide perovskite surfaces remains unknown. This work reveals that electronic states at surfaces of the prototypical CsPbBr3 variant are energetically distributed at room temperature, akin to well-passivated inorganic semiconductors, even when covalent bonds remain cleaved and undercoordinated. Specifically, a striking tendency for shallow surface states is found with approximately 70% of surface-state energies appearing within 0.2 eV or ≈8kBT from the valence-band edge. Furthermore, we show that even when surface states appear deeper in the gap, they are not energetically isolated and are less likely to act as traps. We achieve this result by accelerating first-principles calculations via machine-learning techniques and show that the unique atomic dynamics in these materials render the formation of deep electronic states at their surfaces unlikely. These findings reveal the microscopic mechanism behind the low density of deep defect states at dynamic halide perovskite surfaces, which is key to their exceptional performance in devices. © 2025, CC BY.

关键词： Perovskite

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：