检索结果-内蒙古大学图书馆

Fast Spatiotemporal Learning Framework for Traffic Flow Forecasting

IEEE Transactions on Intelligent Transportation Systems 2023年第8期24卷 8606-8616页

作者： Guo, Canyang Chen, Chi-Hua Hwang, Feng-Jang Chang, Ching-Chun Chang, Chin-Chen Fuzhou University College of Computer and Data Science Fuzhou350108 China National Sun Yat-sen University Department of Business Management Kaohsiung804 Taiwan University of Warwick Department of Computer Science CoventryCV4 7AL United Kingdom Feng Chia University Department of Information Engineering Taichung407802 Taiwan

The graph convolution network (GCN), whose flexible convolution kernels perfectly adapt to the complex topology of the road network, has gradually dominated the spatiotemporal dependency learning of traffic flow data. Defining and learning the spatiotemporal characteristics and relationships of the traffic network efficiently and accurately, which are the important prerequisites for the success of the GCN, have become one of the most burning research problems in the field of intelligent transportation systems. This paper proposes a fast spatiotemporal learning (FSTL) framework containing the fast spatiotemporal GCN module, which reduces the computational complexity of the spatiotemporal GCN from O(k2} to O(k) , where k is the number of time steps of data learned in each GCN operation. To mine globally and fast the correlations of road node pairs, a correlation analysis based on the normal distribution with the complexity of O(N)}, where N is the number of nodes in the traffic network, is proposed to construct the global correlation matrix. Besides, the multi-scale temporal learning is integrated into the FSTL to overcome the receptive field constraints of the spatiotemporal GCN. The experimental results on four real-world datasets demonstrate that the FSTL achieves 48.88% and 5.26% reductions in the training time and mean absolute error, respectively, compared with the state-of-the-art model. © 2000-2011 IEEE.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Guest Editorial: AutoML for Nonstationary Data

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第6期5卷 2456-2457页

作者： Cheng, Ran Escalante, Hugo Jair Tu, Wei-Wei Rijn, Jan N. Van Wang, Shuo Yang, Yun Southern University of Science and Technology Department of Computer Science and Engineering Shenzhen518055 China Instituto Nacional de Astrofísica Óptica y Electrónica Computer Science Department Puebla72840 Mexico 4Paradigm Inc. Beijing100089 China RA Leiden2300 Netherlands The University of Birmingham School of Computer Science BirminghamB15 2TT United Kingdom Yunnan University National Pilot School of Software Kunming650504 China

The five papers in this special section address different aspects of automated machine learning (AutoML) from fundamental algorithms to real-world applications. Developing high-performance machine learning models is a difficult task that usually requires expertise from data scientists and knowledge from domain experts. To make machine learning more accessible and ease the labor-intensive trial-and-error process of searching for the most appropriate machine learning algorithm and the optimal hyperparameter setting, AutoML was developed and has become a rapidly growing area in recent years. AutoML aims at automation and efficiency of the machine learning process across domains and applications. Nowadays, data is commonly collected over time and susceptible to changes, such as in Internet-of-Things (IoT) systems, mobile phone applications and healthcare data analysis. It poses new challenges to the traditional AutoML with the assumption of data stationarity. Interesting research questions arise around whether, when and how to effectively and efficiently deal with non-stationary data in AutoML. © 2020 IEEE.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

The SWH-Analytics Framework 2

The SWH-Analytics Framework

引用

2nd Italian Conference on Big Data and Data science, ITADATA 2023

作者： Antelmi, Alessia Torquati, Massimo Gregori, Daniele Polzella, Francesco Spinatelli, Gianmarco Aldinucci, Marco Computer Science Department University of Torino Italy Computer Science Department University of Pisa Italy HPC-KTT national lab CINI Italy E4 Computer Engineering SpA Italy Zerodivision systems Srl Italy

The Software Heritage (SWH) dataset serves as a vast repository for open-source code, with the ambitious goal of preserving all publicly available open-source projects. Despite being designed to effectively archive project files, its size of nearly 1 petabyte presents challenges in efficiently supporting Big Data MapReduce or AI systems. To address this disparity and enable seamless custom analytics on the SWH dataset, we present the SWH-Analytics (SWHA) architecture. This development environment quickly and transparently runs custom analytic applications on open-source software data preserved over time by SWH. © 2023 Copyright for this paper by its authors.

关键词： Open source software

来源：评论

学校读者我要写书评

暂无评论

Toward Personalization of User Preferences in Partially Observable Smart Home Environments

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2023年第3期4卷 549-561页

作者： Suman, Shashi Rivest, Francois Etemad, Ali Queen's University Department of Electrical and Computer Engineering KingstonONK7L 3N6 Canada Royal Military College Department of Mathematics and Computer Science KingstonONK7K 7B4 Canada

The technologies used in smart homes have recently improved to learn the user preferences from feedback in order to enhance the user convenience and quality of experience. Most smart homes learn a uniform model to represent the thermal preferences of users, which generally fails when the pool of occupants includes people with different sensitivities to temperature, for instance, due to age and physiological factors. Thus, a smart home with a single optimal policy may fail to provide comfort when a new user with a different preference is integrated into the home. In this article, we propose a Bayesian reinforcement learning framework that can approximate the current occupant state in a partially observable smart home environment using its thermal preference and, then, identify the occupant as a new user or someone who is already known to the system. Our proposed framework can be used to identify users based on the temperature and humidity preferences of the occupant when performing different activities to enable personalization and improve comfort. We then compare the proposed framework with a baseline long short-term memory learner that learns the thermal preference of the user from the sequence of actions that it takes. We perform these experiments with up to five simulated human models each based on hierarchical reinforcement learning. The results show that our framework can approximate the belief state of the current user just by its temperature and humidity preferences across different activities with a high degree of accuracy. © 2020 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Advanced dairy cow monitoring: enhanced detection with precision 3D tracking

引用

Multimedia Tools and Applications 2024年 1-30页

作者： Wang, Ranran Li, Yingxiu Yue, Peng Yuan, Chunhong Tian, Fuyang Lu, Xin College of Mechanical and Electronic Engineering Shandong Agricultural University Shandong Tai’an271018 China Department of Computer Science Faculty of Engineering University of Sheffield SheffieldS1 4DP United Kingdom Faculty of Agriculture Iwate University Iwate Morioka020-8550 Japan Faculty of Science and Engineering Iwate University Iwate Morioka020-8550 Japan

Ensuring the welfare of dairy cows requires precise monitoring of their daily exercise to evaluate their physical health. This necessitates innovative methods beyond traditional motion sensors. We present a novel method that integrates an enhanced YOLOv5s object detection model with the DeepSORT multi-object tracking algorithm to meticulously track dairy cow movements, providing holistic information about their health. Our research started with the establishment of a dedicated dataset tailored for cow detection. We then segmented the detection scope to focus on specific regions of interest. Within the modified YOLOv5s model, we replaced the standard CSPDarknet53 backbone with DenseNet to achieve deep separable convolution and feature reorganization modules, leading to reduced parameters, augmented feature expression, and better information flow. In particular, the SPD-Conv module was incorporated to retain intricate details, essential for detecting smaller and low-resolution targets. The transition from Generalized Intersection over Union (GIoU) Loss to Complete Intersection over Union (CIoU) loss improved detection accuracy and sped up model convergence. Our clustering approach, based on the elbow rule, optimized K-means clustering, enhancing speed and precision. For multi-object tracking, the DeepSORT model was tailored to cater to varying cow sizes, and we chose an algorithm to associate appearance information. We converted pixel data into real-world distances, providing exact 3D cow movement metrics. Experimental validation confirmed the efficacy of our approach. Our enhanced model surpassed the original YOLOv5s in performance by 11.1% for accuracy (97.4%), 9.6% for recall (97.8%), and 11.0% for average accuracy (98.2%). The comprehensive accuracy stood at 92.1% for our model. In conclusion, our innovative methodology offers a non-invasive means to monitor dairy cow exercise, paving the way for advanced health assessment techniques in the dairy sector. © The Aut

关键词： K-means clustering

来源：评论

学校读者我要写书评

暂无评论

Error Correction for DNA Storage

IEEE BITS the Information Theory Magazine

引用

IEEE BITS the Information Theory Magazine 2023年第3期3卷 78-94页

作者： Sima, Jin Raviv, Netanel Schwartz, Moshe Bruck, Jehoshua University of Illinois Urbana-Champaign Department of Electrical and Computer Engineering UrbanaIL61801 United States Washington University in Saint Louis Department of Computer Science and Engineering McKelvey School of Engineering St. LouisMO63130 United States McMaster University Department of Electrical and Computer Engineering HamiltonONL8S 4L8 Canada California Institute of Technology Electrical Engineering Department PasadenaCA91125 United States

DNA-based storage is an emerging technology that provides high information density and longevity. Noise and errors are present in almost every stage of the process: writing, storing, and reading. Thus, efficient error correction is crucial to guarantee high reliability and low cost of storing data in DNA. Due to technological constraints and biological limitations, error correction in DNA-based storage poses several coding-theoretic challenges, some of which are new. In this paper, we briefly introduce some of these challenges, including deletion/insertion correcting codes, codes over sliced channels, and duplication-correcting codes. We describe some of the major concepts and solutions in the respective topics, and provide bibliographic notes that briefly review the related literature. © 2021 IEEE.

关键词： DNA Codes Redundancy Genetic communication Symbols Microorganisms Buffer storage Memory Density measurement Noise measurement

来源：评论

学校读者我要写书评

暂无评论

Multiplication of Sparse Matrices and their Transpose Using Compressed Sparse Diagonals

Multiplication of Sparse Matrices and their Transpose Using ...

引用

2024 IEEE High Performance Extreme Computing Conference, HPEC 2024

作者： Haque, Sardar Anisul Parvez, Mohammad Tanvir Hossain, Shahadat Doha Qatar College of Computer Qassim University Department of Computer Engineering Buraydah52571 Saudi Arabia University of Northern British Columbia Department of Computer Science Prince GeorgeBCV2N 4Z9 Canada

ISBN: (纸本)9798350387131

Matrix-matrix multiplication is one of the most important kernel in linear algebra operations with a multitude of applications in scientific and engineering computing. Sparse matrix computation on modern High-performance Computing (HPC) architecture presents challenges such as load balancing, data locality optimization, and computational scalability. Data structures to store sparse matrices are designed to minimize overhead information as well as to optimize the operations count and memory access. In this study, we present a new data structure, 'compressed sparse diagonal' (CSD), to efficiently store and compute with general sparse matrices. The CSD builds upon the previously developed diagonal storage - an orientation-independent uniform scheme to compute with 'structured' matrices [1]. Compared with the widely used compressed sparse row/column (CSR/CSC), the CSD scheme avoids explicit transposition operation when multiplying a matrix with its transpose. The results from preliminary numerical experiments with the aforementioned types of matrices demonstrate the CSD scheme's effectiveness in matrix-transposed matrix multiplication. © 2024 IEEE.

关键词： Matrix algebra

来源：评论

学校读者我要写书评

暂无评论

Empirically evaluating virtual reality’s effect on reservoir engineering tasks

引用

Visual Informatics 2024年第1期8卷 26-46页

作者： Bryson Lawton Nanjia Wang Steven Samoil Parisa Daeijavad Siqi Xie Zhangxin Chen Frank Maurer Department of Computer Science University of Calgary2500 University Dr NWCalgaryT2N 1N4AlbertaCanada Department of Chemical&Petroleum Engineering University of Calgary2500 University Dr NWCalgaryT2N 1N4AlbertaCanada

To help determine in what ways virtual reality(VR)technologies may benefit reservoir engineering workflows,we conducted a usability study on a prototype VR tool for performing reservoir model analysis *** leveraging the strengths of VR technologies,this tool’s aim is to help advance reservoir analysis workflows beyond conventional methods by improving how one understands,analyzes,and interacts with reservoir model *** evaluate our tool’s VR approach to this,the study presented herein was conducted with reservoir engineering experts who used the VR tool to perform three common reservoir model analysis tasks:the spatial filtering of model cells using movable planes,the cross-comparison of multiple models,and well path *** study found that accomplishing these tasks with the VR tool was generally regarded as easier,quicker,more effective,and more intuitive than traditional model analysis software while maintaining a feeling of low task workload on ***,participants provided positive feedback regarding their experience with using VR to perform reservoir engineering work tasks,and in general,it was found to improve multi-model cross-analysis and rough object manipulation in *** indicates the potential for VR to be better than conventional means for some work tasks and participants also expressed they could see it best utilized as an addition to current software in their reservoir model analysis *** were,however,some concerns voiced when considering the full adoption of VR into their work that would be best first addressed before this took place.

关键词： Immersive analytics Visual analytics Data visualization Reservoir engineering Virtual reality VR

来源：评论

学校读者我要写书评

暂无评论

Application of Artificial Bee Colony Algorithm in Solving Second-Order Differential Equations

引用

SN computer science 2024年第8期5卷 1-13页

作者： Murugesh, V. Godla, Sanjiv Rao Meganathan, R. Kumar, G. V. Sam Murugesan, Pandiyanathan Priyadharshini, M. Department of CSE Koneru Lakshmaiah Education Foundation AP Vaddeswaram India Department of Computer Science & amp Engineering Aditya University AP Surampalem India Department of Computing De Montfort University Kazakhstan 120/48 Premises 4 Al-Farabi Ave Medeu District Almaty 050044 Kazakhstan Department of Computer Science and Engineering Faculty of Science and Technology (Icfai Tech) ICFAI Foundations for Higer Education Hyderabad 501 203 India

This paper presents the use of the Artificial Bee Colony (ABC) algorithm to solve second order initial value problems (IVPs). The numerical methods like RK4 and RK-Gill are not very efficient in terms of accuracy, versatility and computational time especially with large step sizes. Based on the foraging behavior of honey bees the ABC approach is capable of enhancing the search for solutions in the search space to achieve the best solution. In this paper we use the second order IVPs to test the performance of the ABC algorithm and compare it with the exact solution, RK4, and RK-Gill methods. The efficacy of the proposed ABC algorithm in terms of accuracy and comparison to other methods is shown in tables and heat maps with zero visible error on several datasets. This shows that the proposed method is accurate and consistent in providing numerical solutions free from rounding off errors. Furthermore, we investigate the combination of the ABC algorithm with conventional numerical techniques and state-of-the-art computing approaches including adaptive step size methods and parallel computing in order to enhance the solution quality and accuracy. The paper also examines how integrating the proposed ABC algorithm with intelligent and machine learning approaches can be useful in extending the possibilities of numerical methods for solving differential equations and demonstrates its versatility as a computational intelligence method for different applications. © The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd. 2024.

关键词： Artificial bee colony (ABC) algorithms and second order initial value problems RK-gill method (RK-Gill) Runge kutta fourth order method (RK4)

来源：评论

学校读者我要写书评

暂无评论

A bio-inspired metaheuristic approach for cloud task scheduling using lateral hyena based particle swarm optimization

引用

Multimedia Tools and Applications 2024年 1-24页

作者： Malik, Meena Nandan, Durgesh Prabha, Chander Uddin, Mueen Acharya, Biswaranjan Hu, Yu-Chen Chandigarh University Mohali India School of CS & amp AI SR University Warangal506371 India Chitkara University Institute of Engineering and Technology Chitakara University Rajpura India College of Computing and IT University of Doha for Science & ampTechnology Doha Qatar Department of Computer Engineering AI & ampBDA Marwadi University Gujarat Rajkot India Department of Computer Science Tunghai University No. 1727 Sec. 4 Boulevard Xitun District Taichung City407224 Taiwan

Being a rapidly emerging field, cloud computing encourages the rapid growth of all the essential internet services. One major challenge in a cloud computing environment is balancing the load among various machines to optimize the appropriate workload, thereby providing efficient task scheduling. Various studies have been proposed and implemented for task scheduling to provide a fair load balance among all the available virtual machines on the cloud. This research work targets the optimization of virtual machines by Lateral Hyena -Particle Swarm Optimization (LH-PSO) algorithm. Our main motivation is to push parallel task scheduling proficiently, producing low response time and rapid results for all the assigned tasks. This work attempts to optimize virtual machines by proposing a Lateral Hyena-based solution and parallel task scheduling. The proposed solution provides a flexible system where virtual machines will be load-balanced by serving equal tasks. The proposed solution attempts to optimize a fitness function based on the velocity and position value of particles mapped to the characteristics of incoming requests. The practice aims to find optimized virtual machines solving tasks with an optimal solution. The performance is analyzed using the Cloudsim simulator to assess all parameters such as load distribution, average response time, memory utilization, processor utilization, and average turnaround time. The analytical conclusions represent an enhancement over the existing schemes compared to the above-mentioned parameters. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Virtual machine

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：