检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Dereich, Steffen Do, Thang Jentzen, Arnulf Weber, Frederic Institute for Mathematical Stochastics Faculty of Mathematics and Computer Science University of Münster Germany School of Data Science The Chinese University of Hong Kong Shenzhen [CUHK-Shenzhen China Deparment of Probability and Statistic Institute of Mathematics Vietnam Academy of Science and Technology Viet Nam School of Data Science Shenzhen Research Institute of Big Data The Chinese University of Hong Kong Shenzhen [CUHK-Shenzhen China Applied Mathematics: Institute for Analysis and Numerics Faculty of Mathematics and Computer Science University of Münster Germany

Deep learning algorithms – typically consisting of a class of deep artificial neural networks (ANNs) trained by a stochastic gradient descent (SGD) optimization method – are nowadays an integral part in many areas of science, industry, and also our day to day life. Roughly speaking, in their most basic form, ANNs can be regarded as functions that consist of a series of compositions of affine-linear functions with multidimensional versions of so-called activation functions. One of the most popular of such activation functions is the rectified linear unit (ReLU) function R ∋ x 7→ max{x, 0} ∈ R. The ReLU function is, however, not differentiable and, typically, this lack of regularity transfers to the cost function of the supervised learning problem under consideration. Regardless of this lack of differentiability issue, deep learning practioners apply SGD methods based on suitably generalized gradients in standard deep learning libraries like TensorFlow or Pytorch. In this work we reveal an accurate and concise mathematical description of such generalized gradients in the training of deep fully-connected feedforward ANNs and we also study the resulting generalized gradient function analytically. Specifically, we provide an appropriate approximation procedure that uniquely describes the generalized gradient function, we prove that the generalized gradients are limiting Fréchet subgradients of the cost functional, and we conclude that the generalized gradients must coincide with the standard gradient of the cost functional on every open sets on which the cost functional is continuously *** Codes 68T07 Copyright © 2025, The Authors. All rights reserved.

关键词： Cost functions

来源：评论

学校读者我要写书评

暂无评论

A Unique Method of Detection of Edges and Circles of Multiple Objects in Imaging Scenarios Using Line Descriptor Concepts 1

引用

International Conference on Advances in IoT, Security with AI, ICAISA 2023

作者： Patil, Rajesh Maharudra Nagaraja, B.G. Prasad, M.R. Manjunath, T.C. Rayappa, Ravi Electrical Engineering Department SKNS College of Engineering Korti Affiliated to Solapur University Maharashtra Pandharpur India Electronics and Communication Engineering Vidyavardhaka College of Engineering Karnataka Mysuru India Computer Science and Engineering Vidyavardhaka College of Engineering Karnataka Mysuru India Electronics and Communication Engineering Deparment Dayananda Sagar College of Engineering Bengaluru India Electronics and Communication Engineering Jain Institute of Technology Karnataka Davanagere India

ISBN: (数字)9789819950881

ISBN: (纸本)9789819950874

This paper proposes a novel method for detecting the edges and circles of multiple objects in imaging scenarios using line descriptor concepts. The method involves analyzing the intensity gradient and the orientation of the pixels in the image, and using this information to identify lines and circles that are likely to correspond to object boundaries. The proposed approach is compared with existing methods and is shown to provide superior performance in terms of accuracy and computational efficiency. The method is particularly useful for applications such as object recognition and tracking, where accurate detection of object boundaries is essential. The experimental results demonstrate the effectiveness of the proposed method in detecting edges and circles of multiple objects in different imaging scenarios. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Object recognition

来源：评论

学校读者我要写书评

暂无评论

Geo-SID: A new geocast safety information dissemination protocol in VANET for urban areas 2

Geo-SID: A new geocast safety information dissemination prot...

引用

2nd International Conference on Computation, Automation and Knowledge Management, ICCAKM 2021

作者： Senouci, Oussama Aliouat, Zibouda Harous, Saad Mohamed EI-Bachir EI-Ibrahimi University Computer Science Deparment Bordj Bou arreridj Algeria LRSD Laboratory Ferhat Abbas Setif1 University Computer Science Dept. Setif Algeria College of Information Technology University Al-Ain United Arab Emirates

ISBN: (纸本)9781728194912

Recently, many researches on Vehicular Ad Hoc Networks (VANETs) have been performed because of the many related applications. Data dissemination is among the challenging issues face by these kinds of networks to overcome several problems, such as broadcast storm, and network fragmentation. In this respect, the present work proposes a new Geocast-based Safety Information Dissemination protocol for VANETs, called Geo-SID. The proposed protocol is used for dissemination of safety information in a V2V scenario. It is specifically designed and suitable for an urban environment. The main goal of Geo-SID scheme is to perform the safety information dissemination process within a Zone of Relevance (ZoR) and without using the Zone of Forwarding (ZoF) similar to existing dissemination protocols in the literature. Furthermore, our solution aims to eliminate the broadcast storm problem by avoiding irrelevant alert messages and providing an efficient safety data dissemination across a ZoR area by extending the coverage area, reducing the network overhead and improving the latency. We simulate our proposed protocol using network simulator NS-3 and SUMO environment. Our proposed protocol outperforms similar protocols available in the literature based on the obtained simulation results. © 2021 IEEE.

关键词： Vehicular ad hoc networks

来源：评论

学校读者我要写书评

暂无评论

Speech Denoising Based on Wavelet Transform and Wiener Filtering

Speech Denoising Based on Wavelet Transform and Wiener Filte...

引用

International Conference on Frontiers of Signal Processing (ICFSP)

作者： Marjan Joorabchi Seyed Ghorshi Yashar Naderahmadian School of Science and Engineering Sharif University of Technology – International Campus Kish Island Iran Department of Electrical and Computer Engineering The University of Texas at Tyler Tyler Texas USA Deparment of Electrical Engineering University of Guilan Rasht Iran

This paper presents a combinational method of Wavelet Transform and Adaptive Filter to reduce distortion that has been caused by interference noise. The procedure makes use of a Wiener filtering to improve and enhance the quality of the speech signal. The proposed method is tested for a couple of noisy speech signals with different SNRs in the range of −5 dB to 15 dB. In this method we aim to inspect the enhancement of noisy speech signals in more practical conditions. Thus, not only is the process employed on noisy speech signals distorted by Additive White Gaussian Noise (WGN) but also it is practiced on environmental noises such as CAR noise, etc. The results show improvement of noise reduction after applying Wiener filtering to the Wavelet Transform compared with utilizing Wavelet Transform only.

关键词：

来源：评论

学校读者我要写书评

暂无评论

PheME: A deep ensemble framework for improving phenotype prediction from multi-modal data

arXiv

引用

arXiv 2023年

作者： Zhang, Shenghan Li, Haoxuan Tang, Ruixiang Ding, Sirui Rasmy, Laila Zhi, Degui Zou, Na Hu, Xia Department of Computer Science Xi’an Jiaotong-Liverpool University Xi’an China Department of Computer Science Wuhan University Wuhan China Department of Computer Science Rice University Houston United States Department of Computer Science and Engineering Texas A&M University College Station United States UTHealth Houston School of Biomedical Informatics Houston United States Department of Engineering Technology & Industrial Distribution Texas A&M Univerisity College Station United States Deparment of Computer Science Rice University Houston United States

Detailed phenotype information is fundamental to accurate diagnosis and risk estimation of diseases. As a rich source of phenotype information, electronic health records (EHRs) promise to empower diagnostic variant interpretation. However, how to accurately and efficiently extract phenotypes from the heterogeneous EHR data remains a challenge. In this work, we present PheME, an Ensemble framework using Multi-modality data of structured EHRs and unstructured clinical notes for accurate Phenotype prediction. Firstly, we employ multiple deep neural networks to learn reliable representations from the sparse structured EHR data and redundant clinical notes. A multi-modal model then aligns multi-modal features onto the same latent space to predict phenotypes. Secondly, we leverage ensemble learning to combine outputs from single-modal models and multi-modal models to improve phenotype predictions. We chose seven diseases to evaluate the phenotyping performance of the proposed framework. Experimental results show that using multi-modal data significantly improves phenotype prediction in all diseases, the proposed ensemble learning framework can further boost the performance. Copyright © 2023, The Authors. All rights reserved.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Future Food Production Prediction Using AROA Based Hybrid Deep Learning Model in Agri-Sector

引用

Human-Centric Intelligent Systems 2023年第4期3卷 521-536页

作者： Baswaraju, Swathi Maheswari, V. Uma Chennam, krishna Keerthi Thirumalraj, Arunadevi Kantipudi, M. V. V. Prasad Aluvalu, Rajanikanth Deparment of CSE New Horizon College of Engineering Bengaluru India Department of CSE Chaitanya Bharathi Institute of Technology Hyderabad India Department of CSE Vasavi College of Engineering Hyderabad India Department of Computer Science Engineering K.Ramakrishnan College of Technology Trichy India Symbiosis International (Deemed University) Symbiosis Institute of Technology Pune India Department of IT Chaitanya Bharathi Institute of Technology Hyderabad India

Policymaking and administration of national tactics of action for food security rely heavily on advances in models for accurate estimation of food output. In several fields, including food science and engineering, machine learning (ML) has been established to be an effective tool for data investigation and modelling. There has been a rise in recent years in the application of ML models to the tracking and forecasting of food safety. In our analysis, we focused on two sources of food production: livestock production and agricultural production. Livestock production was measured in terms of yield, number of animals, and sum of animals slaughtered; crop output was measured in terms of yields and losses. An innovative hybrid deep learning model is proposed in this paper by fusing a Dense Convolutional Network (DenseNet) with a Long Short-Term Memory (LSTM) to do production analysis. The hybridised algorithm, or A-ROA for short, combines the Arithmetic Optimisation Algorithm (AOA) and the Rider Optimisation Algorithm (ROA) to determine the ideal weight of the LSTM. The current investigation focuses on Iran as a case study. Therefore, we have collected FAOSTAT time series data on livestock and farming outputs in Iran from 1961 to 2017. Findings from this study can help policymakers plan for future generations' food safety and supply by providing a model to anticipate the upcoming food construction.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Fast Whole Slide Image Analysis Of Cervical Cancer Using Weak Annotation

Fast Whole Slide Image Analysis Of Cervical Cancer Using Wea...

引用

IEEE International Symposium on Biomedical Imaging

作者： Min Ling Guofeng Lv Jue Wang Xiaoyu Hao Jun Shi Hong An School of Computer Science and Technology University of Science and Technology of China Pathology Deparment The First Affiliated Hospital of USTC West District

ISBN: (纸本)9781665412469;9781665429474

Cervical cancer is the most common gynecological disease, which seriously endangers women's health. Histopathological diagnosis is the most reliable determinant for cervical cancer prognosis and treatment. However, it is a time-consuming and error-prone task for pathologists to manually analyze the whole slide image(WSI). Automated detection of cancer tissue and histopathological diagnosis are of great signification to reduce the workload for pathologists. So we proposed a method combining deep learning(DL) and traditional machine learning(ML) to realize the rapid analysis of the cervical cancer WSI. Our method first applied a patch-level network to detect the pathological patches and generate a heatmap that can highlight the tumor area. Considering the difficulty of obtaining pixel-level annotations, we designed a weak annotation strategy for training the network and verified the feasibility of this strategy on a public dataset. Besides, we used overlapping heatmap generation strategy(Overlapping HMGS) and feature sharing strategy of fully convolutional neural networks(CNN) to balance the accuracy and speed of tumor region detection. Then features extracted from the heatmaps are fed into a traditional machine learning model for the WSI-level prediction. Our approach not only achieved excellent performance with 97% accuracy on our cervical cancer WSI dataset, but it is also efficient and interpretable.

关键词： Heating systems Training Annotations Feature extraction Classification algorithms Reliability Task analysis

来源：评论

学校读者我要写书评

暂无评论

Computational study of the role of counterions and solvent dielectric in determining the conductance of B-DNA

引用

Physical Review E 2023年第4期107卷 044404-044404页

作者： Yiren Wang Busra Demir Hashem Mohammad Ersin Emre Oren M. P. Anantram Deparment of Electrical and Computer Engineering University of Washington Seattle Washington 98105 USA Bionanodesign Laboratory Department of Biomedical Engineering and Department of Materials Science & Nanotechnology Engineering TOBB University of Economics and Technology Ankara 06510 Turkey Department of Electrical Engineering Kuwait University P.O. Box 5969 Safat 13060 Kuwait

DNA naturally exists in a solvent environment, comprising water and salt molecules such as sodium, potassium, magnesium, etc. Along with the sequence, the solvent conditions become a vital factor determining DNA structure and thus its conductance. Over the last two decades, researchers have measured DNA conductivity both in hydrated and almost dry (dehydrated) conditions. However, due to experimental limitations (the precise control of the environment), it is very difficult to analyze the conductance results in terms of individual contributions of the environment. Therefore, modeling studies can help us to gain a valuable understanding of various factors playing a role in charge transport phenomena. DNA naturally has negative charges located at the phosphate groups in the backbone, which provides both the connections between the base pairs and the structural support for the double helix. Positively charged ions such as the sodium ion (Na+), one of the most commonly used counterions, balance the negative charges at the backbone. This modeling study investigates the role of counterions both with and without the solvent (water) environment in charge transport through double-stranded DNA. Our computational experiments show that in dry DNA, the presence of counterions affects electron transmission at the lowest unoccupied molecular orbital energies. However, in solution, the counterions have a negligible role in transmission. Using the polarizable continuum model calculations, we demonstrate that the transmission is significantly higher at both the highest occupied and lowest unoccupied molecular orbital energies in a water environment as opposed to in a dry one. Moreover, calculations also show that the energy levels of neighboring bases are more closely aligned to ease electron flow in the solution.

关键词： Electrical conductivity Electronic structure of atoms & molecules Quantum transport DNA Nucleic acids Density functional theory Methods in transport Nonequilibrium Green's function

来源：评论

学校读者我要写书评

暂无评论

Retraction Note: Enabling secure and efficient industry 4.0 transformation through trust-authorized anomaly detection in cloud environments with a hybrid AI approach

引用

Optical and Quantum Electronics 2024年第10期56卷 1-2页

作者： Prakash, N. Vignesh, J. Ashwin, M. Ramadass, Sudhir Veeranjaneyulu, N. V. Athawale, Shashikant Ravuri, Ananda Subramanian, Balambigai Department of Information Technology B.S Abdur Rahman Crescent Institute of Science and Technology Vandalur India Department of ISE Jain University Bengaluru India Deparment of Artificial Intelligence and Data Science Koneru Lakshmaiah Education Foundation Vaddeswaram India Sterck Systems Chennai India Department of Information Technology VFSTR Deemed to Be University Vadlamudi Guntur District India Department of Computer Engineering AISSMS COE Pune India Intel Corporation Hillsboro USA Department of ECE Kongu Engineering College Perundurai India

来源：评论

学校读者我要写书评

暂无评论

Android based Recommender System (ARS) to Detect Breast Abnormalities

Android based Recommender System (ARS) to Detect Breast Abno...

引用

Automation, Computing and Renewable Systems (ICACRS), International Conference on

作者： M Kavitha P Venkata Krishna V Rama Krishna Sreenivasulu Digavinti M Kalyani Tiruluka V S Naga Divya Department of Computer Science and Engineering Koneru Lakshmaiah Education Foundation Vaddeswaram AP India Deparment of Computer Science Sri Padmavathi Mahila Visvavidyalam Tirupati India Vignan’s Foundation for Science Technology and Research Vadlamudi India Department of Computer Science & Engineering (Data Science) Institute of Aeronautical Engineering Hyderabad India Department of Mathematics PACE institute of technology and Sciences Ongole AP India

ISBN: (纸本)9781665460859

Breast tumor is the second leading cause of death among women from the last six decades. The researchers are working intensely on identifying breast tumor symptoms at an early stage. Early detection and exact treatment of tumors will support expanding the lifetime of a patient. Doctors recommend that periodic Breast-Self Examination (BSE) is the preliminary breast cancer screening test. This test will assist the people to detect breast abnormalities at an early stage. But due to a lack of awareness and knowledge on the test procedure of BSE, people are unable to take this self-test periodically. An Android based Recommender System (ARS) to detect breast abnormalities are proposed to serve this gap. ARS is an android based mobile app developed to support all categories of people in breast health monitoring. Through this recommender system, the breast physical symptoms are assessed, and the analysis report helps the user to detect the abnormality at an early stage.

关键词： Renewable energy sources Breast tumors Medical services Breast Built-in self-test Breast cancer Mobile applications

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：