检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Mehra, Anket Prieß, Malte Himstedt, Marian Department of Computer Science and Electrical Engineering University of Applied Sciences Kiel Kiel Germany Department of Electrical Engineering and Computer Science Technical University of Applied Sciences Lübeck Lübeck Germany

This paper shows that further evaluation metrics during model training are needed to decide about its applicability in inference. As an example, a LayoutLM-based model is trained for token classification in documents. The documents are German receipts. We show that conventional classification metrics, represented by the F1-Score in our experiments, are insufficient for evaluating the applicability of machine learning models in practice. To address this problem, we introduce a novel metric, Document Integrity Precision (DIP), as a solution for visual document understanding and the token classification task. To the best of our knowledge, nothing comparable has been introduced in this context. DIP is a rigorous metric, describing how many documents of the test dataset require manual interventions. It enables AI researchers and software developers to conduct an in-depth investigation of the level of process automation in business software. In order to validate DIP, we conduct experiments with our created models to highlight and analyze the impact and relevance of DIP to evaluate if the model should be deployed or not in different training settings. Our results demonstrate that existing metrics barely change for isolated model impairments, whereas DIP indicates that the model requires substantial human interventions in deployment. The larger the set of entities being predicted, the less sensitive conventional metrics are, entailing poor automation quality. DIP, in contrast, remains a single value to be interpreted for entire entity sets. This highlights the importance of having metrics that focus on the business task for model training in production. Since DIP is created for the token classification task, more research is needed to find suitable metrics for other training tasks. Copyright © 2025, The Authors. All rights reserved.

关键词： Natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

Data-driven Error Estimation: Upper Bounding Multiple Errors without Class Complexity as Input

arXiv

引用

arXiv 2024年

作者： Krishnamurthy, Sanath Kumar Lyubarskaja, Anna Brunskill, Emma Athey, Susan Meta Central Applied Science United States Institute for Computational and Mathematical Engineering Stanford University United States Department of Computer Science Stanford University United States Stanford University Graduate School of Business United States Stanford Institute for Human-Centered Artificial Intelligence United States

Constructing confidence intervals that are simultaneously valid across a class of estimates is central for tasks such as multiple mean estimation, bounding generalization error in machine learning, and adaptive experimental design. We frame this as an "error estimation problem," where the goal is to determine a high-probability upper bound on the maximum error for a class of estimates. We propose an entirely data-driven approach that derives such bounds for both finite and infinite class settings, naturally adapting to a potentially unknown correlation structure of random errors. Notably, our method does not require class complexity as an input, overcoming a major limitation of existing approaches such as union bounding and bounds based on Talagrand’s inequality. In this paper, we present our simple yet general solution and demonstrate its flexibility through applications ranging from constructing multiple simultaneously valid confidence intervals to optimizing exploration in contextual bandit algorithms. © 2024, CC BY-NC-ND.

关键词： Random errors

来源：评论

学校读者我要写书评

暂无评论

Radial kinks in a Schwarzschild-like geometry

引用

Physical Review D 2024年第12期110卷 125025-125025页

作者： Jean-Guy Caputo Tomasz Dobrowolski Jacek Gatlik Panayotis G. Kevrekidis Laboratoire de Mathématiques INSA de Rouen Avenue de l’Université 76801 Saint-Etienne du Rouvray France Department of Computer Physics and Quantum Computing University of the National Education Commission in Krakow Podchora¸żych 2 30-084 Cracow Poland AGH University of Krakow Faculty of Physics and Applied Computer Science 30-059 Krakow Poland Department of Mathematics and Statistics University of Massachusetts Amherst Massachusetts 01003-4515 USA

We study the propagation of a domain wall (kink) of the ϕ4 model in a radially symmetric environment defined by a gravity source. This source introduces a Schwarzschild-like geometry. We introduce an effective model that accurately describes the dynamics of the kink center. This description works well even outside the perturbation region, i.e., even for large masses of the gravitating object. We observed that such a spherical domain wall surrounding a star-type object inevitably “collapses,” i.e., shrinks in radius toward the origin and offer an understanding of the latter phenomenology. The relevant analysis is presented for a circular domain wall and a spherical one.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Local negative permittivity and topological phase transition in polar skyrmions (vol 20, pg 194, 2021)

引用

NATURE MATERIALS 2021年第6期20卷 905-905页

作者： Das, S. Hong, Z. Stoica, V. A. Goncalves, M. A. P. Shao, Y. T. Parsonnet, E. Marksz, E. J. Saremi, S. McCarter, M. R. Reynoso, A. Long, C. J. Hagerstrom, A. M. Meyers, D. Ravi, V. Prasad, B. Zhou, H. Zhang, Z. Wen, H. Gomez-Ortiz, F. Garcia-Fernandez, P. Bokor, J. iniguez, J. Freeland, J. W. Orloff, N. D. Junquera, J. Chen, L. Q. Salahuddin, S. Muller, D. A. Martin, L. W. Ramesh, R. Department of Materials Science and Engineering University of California Berkeley CA USA Department of Physics University of California Berkeley CA USA Department of Materials Science and Engineering The Pennsylvania State University University Park PA USA Department of Mechanical Engineering Carnegie Mellon University Pittsburgh PA USA Materials Research and Technology Department Luxembourg Institute of Science and Technology (LIST) Esch/Alzette Luxemburg Departamento de Ciencias de la Tierra y Física de la Materia Condensada Universidad de Cantabria Cantabria Campus Internacional Santander Spain Physics and Materials Science Research Unit University of Luxembourg Belvaux Luxembourg School of Applied and Engineering Physics Cornell University Ithaca NY USA National Institute of Standards and Technology Boulder CO USA Advanced Photon Source Argonne National Laboratory Argonne IL USA Department of Electrical Engineering and Computer Sciences University of California Berkeley CA USA Kavli Institute at Cornell for Nanoscale Science Ithaca NY USA Materials Sciences Division Lawrence Berkeley National Laboratory Berkeley CA USA

A Correction to this paper has been published: https://***/10.1038/s41563-021-00962-z.

关键词： Ferroelectrics and multiferroics

来源：评论

学校读者我要写书评

暂无评论

Comparison of the Effectiveness of Machine Learning Techniques for the Number of Road Traffic Injuries

Comparison of the Effectiveness of Machine Learning Techniqu...

引用

International ECTI Northern Section Conference on Electrical, Electronics, computer and Telecommunications engineering (ECTI-NCON)

作者： Wirawan Puttamat Wikanda Phaphan Anuchit Jitpattanakul Chainarong Pianpailoon Sakorn Mekruksavanich Department of Mathematics Faculty of Education Chaiyaphum Rajabhat University Chaiyaphum Thailand Department of Applied Statistics Faculty of Applied Science King Mongkut’s University of Technology North Bangkok Bangkok Thailand Research Group in Statistical Learning and Inference Department of Mathematics Faculty of Applied Science King Mongkut’s University of Technology North Bangkok Bangkok Thailand Intelligent and Nonlinear Dynamic Innovations Research Center Curriculum and Teaching Program (Mathematics) Faculty of Education Sakon Nakhon Rajabhat University Sakon Nakhon Thailand Department of Computer Engineering School of Information and Communication University of Phayao Phayao Thailand

This article aims to compare forecasting techniques with machine learning techniques for predicting the number of people injured in road accidents using data from the Injury Information Collaboration Center, department of Disease Control, Ministry of Public Health Outpatient Files (OPD) spanning the years 2018 to 2022. The four machine-learning techniques examined in this study include Decision Tree Regression, Random Forest Regression, Support Vector Regression, and Multiple Linear Regression. The data analysis was performed using the R programming language. The results revealed that the Decision Tree Regression technique yielded the most accurate predictions for the number of road traffic injuries, as evidenced by its lowest values of MAE, MSE, and RMSE.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Analyzing Generalized Pólya Urn Models Using Martingales, with an Application to Viral Evolution

SSRN

引用

SSRN 2023年

作者： Specht, Ivan Mitzenmacher, Michael Harvard College Faculty of Arts and Sciences Harvard University CambridgeMA02138 United States Department of Computer Science School of Engineering and Applied Sciences Harvard University CambridgeMA02138 United States

The randomized play-the-winner (RPW) model is a generalized Pólya Urn process with broad applications ranging from viral genomics to clinical trials. We derive an exact expression for the variance of the RPW model by transforming the Pólya Urn process into a martingale, correcting an earlier result of Matthews and Rosenberger (1997). We then use this result to approximate the full probability mass function of the RPW model for certain parameter values relevant to genetic applications. Finally, we fit our model to genomic sequencing data of SARS-CoV-2, demonstrating a novel method of estimating the viral mutation rate that delivers comparable results to existing scientific literature. © 2023, The Authors. All rights reserved.

关键词： Genome

来源：评论

学校读者我要写书评

暂无评论

Xeggora: Exploiting immune-to-evidence symmetries with full aggregation in statistical relational models 29

Xeggora: Exploiting immune-to-evidence symmetries with full ...

引用

29th International Joint Conference on Artificial Intelligence, IJCAI 2020

作者： Amirian, Mohammad Mahdi Ghidary, Saeed Shiry Department of Computer Engineering Amirkabir University of Technology Tehran Iran Department of Math and Computer Science Amirkabir University of Technology Tehran Iran

ISBN: (纸本)9780999241165

We present improvements in maximum a-posteriori inference for Markov Logic, a widely used SRL formalism. Several approaches, including Cutting Plane Aggregation (CPA), perform inference through translation to Integer Linear Programs. Aggregation exploits context-specific symmetries independently of evidence and reduces the size of the program. We illustrate much more symmetries occurring in long ground clauses that are ignored by CPA and can be exploited by higher-order aggregations. We propose Full-Constraint-Aggregation, a superior algorithm to CPA which exploits the ignored symmetries via a lifted translation method and some constraint relaxations. RDBMS and heuristic techniques are involved to improve the overall performance. We introduce Xeggora as an evolutionary extension of RockIt, the query engine that uses CPA. Xeggora evaluation on real-world benchmarks shows progress in efficiency compared to RockIt especially for models with long formulas. © 2020 Inst. Sci. inf., Univ. Defence in Belgrade. All rights reserved.

关键词： Integer programming

来源：评论

学校读者我要写书评

暂无评论

A structure-activity relationship modelling of opioid compounds by using machine learning 15

A structure-activity relationship modelling of opioid compou...

引用

15th Conference of the Euro-American Consortium for Promoting the Application of mathematics in Technical and Natural sciences, AMiTaNS 2023

作者： Sapundzhi, Fatima Lazarova, Meglena Dzimbova, Tatyana Georgiev, Slavi Ivanova, Antonina Department of Communication and Computer Engineering Faculty of Engineering South-West University Neofit Rilski 66 Ivan Myhailov Str. Blagoevgrad2700 Bulgaria Faculty of Applied Mathematics and Informatics Technical University of Sofia 8 St. Kliment Ohridski Blvd. Sofia1000 Bulgaria Institute of Molecular Biology Bulgarian Academy of Sciences Acad. G.Bonchev Str bl. 21 Sofia1113 Bulgaria Department of Information Modeling Institute of Mathematics and Informatics Bulgarian Academy of Sciences Acad. Georgi Bonchev Str. bl. 8 Sofia1113 Bulgaria Department of Applied Mathematics and Statistics Faculty of Natural Sciences and Education University of Ruse 8 Studentska Str. Ruse7004 Bulgaria Department of Computer Science Faculty of International Economics and Administration Varna Free University Chernorizets Hrabar 84 Yanko Slavchev Str. Chaika Resort Varna9007 Bulgaria

Opiates are among the oldest drugs that are used to treat many medical problems. They are analgesic and sedative drugs that contain opium. The morphine is its most active ingredient and it is a widely used pain reliever despite its side effects. The main objective of this study is to construct a model which gives the structure-activity relationship among a series of mu-opioid ligands and molecular docking results. For this purpose, a model of mu-opioid receptors using machine learning is introduced. By obtaining a relationship between the docking results and the in vivo test, we could predict the biological effect of the newly synthesized ligands. © Published under licence by IOP Publishing Ltd.

关键词： Medical problems

来源：评论

学校读者我要写书评

暂无评论

In Situ Formed Bimetal Metal-Organic Frameworks in Polyvinyl Alcohol/Sodium Alginate-Based Conductive Hydrogels Towards Soft Electronics

SSRN

引用

SSRN 2024年

作者： Li, Jiongru Wei, Huige Cui, Shuaichuan Hou, Hua Zhang, Yifan Zhang, Yingying Xu, Ben Bin El-Bahy, Zeinhom M. Melhi, Saad Sellami, Rahma Guo, Zhanhu Tianjin Key Laboratory of Brine Chemical Engineering and Resource Eco-utilization College of Chemical Engineering and Materials Science Tianjin University of Science and Technology Tianjin300457 China State Key Laboratory of Polymer Physics and Chemistry Changchun Institute of Applied Chemistry Chinese Academy of Sciences Changchun130022 China College of Materials Science and Engineering Taiyuan University of Science and Technology Taiyuan030024 China Tianjin University of Science and Technology China Department of Mechanical and Construction Engineering Northumbria University Newcastle Upon TyneNE1 8ST United Kingdom Department of Chemistry Faculty of Science Al-Azhar University Nasr City Cairo11884 Egypt Department of Computer Science Applied College Northern Border University Rafha91911 Saudi Arabia University of Bisha Saudi Arabia

Zeolitic imidazolate frameworks (ZIFs) have received enormous attention due to unique physi-chemical properties, but are rarely reported for applications in electrically conductive hydrogels (ECHs) arising from low intrinsic conductivity and poor dispersion. Herein, we propose an innovative strategy to prepare highly conductive and mechanically robust ECHs by in situ growing Ni/Co bimetallic ZIFs within the polyvinyl alcohol/sodium alginate dual network. 2-methylimidazole (MeIM) ligands are creatively copolymerized with pyrrole monomers to form Py-co-MeIM conductive fillers and improve the conjugation degree of the polymer chains and thereby increase the conductivity of the hydrogel. Meanwhile, MeIM ligands act as anchor points for the in-situ formation of bimetallic ZIFs by trapping targeted metal ions, effectively avoiding phase-to-phase interfacial resistance and ensuring a uniform distribution in the hydrogel network. Due to the synergism of Ni/Co bimetallic ZIFs, the PZPS hydrogel exhibits a high areal capacitance, promising for flexible energy storage devices. In addition, the hydrogel shows excellent mechanical strength and toughness, a high gauge factor and a pressure sensitivity over an extremely wide stress range when used as flexible wearable strain/pressure sensors. This study provides new insights to incorporating highly conductive and uniformly dispersed ZIFs into hydrogels for flexible wearable electronics. © 2024, The Authors. All rights reserved.

关键词： Hydrogels

来源：评论

学校读者我要写书评

暂无评论

An Overall Disease Analysis of Mango Using Neural Network with Hybrid Feature model

An Overall Disease Analysis of Mango Using Neural Network wi...

引用

applied Intelligence and Computing (AIC), IEEE World Conference on

作者： Jatinder Kaur Ashwani K Sharma Manudeep Kaushal Arti Badhoutiya Nirmala Reddy Ahmed Alkhayyat Department of ECE Chandigarh Engineering College Jhanjeri Mohali India Department of Applied Sciences Chandigarh Engineering College Jhanjeri Mohali India Dean Professional Studies and Dean Corporate Affairs Gulzar Group of Institutes Khanna Punjab India Department of Electrical Engineering GLA University Mathura India Department of Computer Science & Engineering IES College of Technology Bhopal IES University Bhopal India College of Technical Engineering The Islamic University Najaf Iraq

Mango is a highly valued fruit crop, with arable land covering about 36% of the total fruit area. With roughly 23% of the total area under mango, Uttar Pradesh and Andhra Pradesh lead the way, followed by Tamil Nadu, Karnataka, Gujarat, and Bihar. Global food security is affected a lot due to crop plant diseases. These diseases directly affect the quality of food, fruits, etc., resulting in a fall in agricultural productivity. Farmers must assess the leaf, fruit, or tree through visual interpretation to determine whether any fruit crop part is contaminated or blooming properly. But this traditional technique has its limitations; it is inconsistent, unreliable, and prone to mistakes. Researchers have proposed a variety of techniques for resolving the above-stated issues. Most of them preferred the modelling of convolutional neural network models according to problem areas by giving low-resolution images as input, resulting in low disease detection accuracy. In this work, we use an artificial neural network (ANN) technique with the objective of early detection of disease in any part of the fruit crop with tiny disease blobs that can only be seen with better resolution images. All the contaminated blobs are segmented for the entire dataset after a pre-processing phase using a contrast enhancement approach.

关键词：

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：