检索结果-内蒙古大学图书馆

Formal Modeling and Analysis of Transaction Scheduling and Concurrency Control in Grid Databases

Formal Modeling and Analysis of Transaction Scheduling and C...

Asia Simulation conference 2008/the 7th international conference on System Simulation and Scientific Computing（ICSC‘2008）

作者： Yaojun Han College of international Business,Shanghai international Studies University,Shanghai,200083 CHINA Changjun Jiang Department of Computer Science & Engineering,Tongji University,Shanghai,200092,CHINA Xumei Luo College of international Finance and Trade,Shanghai international Studies University,Shanghai,200083 CHINA

Transaction scheduling and concurrency control are two important problems in a distributed environment. A lot of research has been focused on the problems recently, and many models and measures have been proposed. However, some of them didn't fit for grid environment, as grid architecture is a new and distinguished from conventional distributed computing by its heterogeneous, dynamic and autonomous properties. In this paper, we propose a three-level transaction scheduling model for grid database, which not only shows the autonomy of grid but also lightens the pressure of communication and grid scheduler. We consider communication time of a transaction, user's priority and the number of DB sites accessed by the transaction as a priority of the transaction, and give scheduling algorithms for concurrency transaction, which not only supports correctness of the data, but also improves the effectiveness of the system. Petri nets are promising tools for modeling and analysis information processing systems that are characterized as being concurrent, parallel and distributed. We use a dynamic time colored Petri net (DTCPN) to model each level's transaction scheduling and get a DTCPN model for three-level transaction scheduling model of grid database using sharing composition. Finally, we get some important results such as miss rate of transactions by analyzing reachability of Petri nets.

关键词： net Petri Formal Modeling and Analysis of Transaction Scheduling and Concurrency Control in Grid Databases

来源：评论

学校读者我要写书评

暂无评论

Exploiting Combination of Spectral Features in Light Weight Neural Network for ABE on Limited Data

Exploiting Combination of Spectral Features in Light Weight ...

引用

international conference on Signal processing and Information Security (ICSPIS)

作者： Ritujoy Biswas Karan Nathwani Mohammed Krini Electrical Engineering Indian Institute of Technology Jammu Jammu India Laboratory for Signal Processing Aschaffenburg University of Applied Sciences Aschaffenburg Germany

ISBN: (数字)9798350368673

ISBN: (纸本)9798350368680

this paper proposes an approach to extend the bandwidth of 4 KHz speech signals upto 8 KHz using a simple feed-forward neural network (FFNN). We employ the commonly used Short-Time Fourier Transform (STFT) values as features and boost its resolution by incorporating Mel Frequency Cepstral Coefficients (MFCCs). Apart from this feature-level fusion, a layer-level fusion is also examined to demonstrate the expansion in bandwidth. this is an advantage over traditional artificial bandwidth extension (ABE) methods using neural nets, which are heavily data-driven and require high model complexity. the proposed approach exploits the combination of relevant spectral features to avoid the requirement of highly complex architectures. this is because the combined features capture more intricate features relating the 4 KHz signal to the 8 KHz one. this facilitates the deployment of the technique on resourceconstrained devices, even with limited data. these reconstructed high pass (8 KHz) audio signals are perceptually better than the original 4 KHz bandwidth signals. Additionally, the performance of the proposed approach exceeds that of state-of-the-art methods in the literature. this approach has potential applications in enhancing the quality by extending the bandwidth of audio signals in telecommunication systems and also finds application in cochlear implant (CI) users.

关键词： Performance evaluation Cochlear implants Fourier transforms Information security Bandwidth Telecommunications Feedforward neural networks Complexity theory Mel frequency cepstral coefficient Signal resolution

来源：评论

学校读者我要写书评

暂无评论

Performance Evaluation of Deep Learning Models for Real-Time Trash Classification on Embedded Devices

Performance Evaluation of Deep Learning Models for Real-Time...

引用

international conference on Signal processing and Information Security (ICSPIS)

作者： Abduselam Ahmed Fanuel Petros Tomas Yemane Farid Flitti Yassine Benachour Faculty of Engineering Higher Colleges of Technology Fujairah UAE

ISBN: (数字)9798350368673

ISBN: (纸本)9798350368680

Trash classification represents a significant challenge in waste recycling industry, leading often to contamination due to manual misclassification of recyclable and non-recyclable items. this arises because human judgment relies on experience and knowledge rather than on objective criteria. To address this issue many scholars have proposed automated classification systems based on machine learning and deep learning (DL). All these works have not been tested on real time embedded systems. this paper proposes a comparison between several DL models to categorize waste into cardboard, plastic, glass, metal, paper, and general trash. Several DL architectures were retrained and implemented on a Raspberry Pi, i.e., MobileNet, ResNet, EfficientNet, and DenseNet models. the performance, i.e., classification accuracy, inference time and maximum number of frames per second (FPS) of these models were evaluated using the TrashNet dataset [4]. the obtained results demonstrate that EfficientNet B3 achieved the best performance based on compared parameters as well as a testing accuracy at about $86 \%$, while DenseNet achieved the lowest performance parameters with a low accuracy of $79 \%$. Additionally, the models were evaluated for their performance on embedded devices. According to the experiment conducted in this work, MobileNetV2 was the fastest at classification, while EfficientNet B7 was the slowest. Considering both accuracy and speed, EfficientNet B3 is the best model for real time applications.

关键词： Performance evaluation Deep learning Training Solid modeling Accuracy Signal processing Real-time systems Recycling Plastics Testing

来源：评论

学校读者我要写书评

暂无评论

parallel adaptive estimation of hip range of motion for total hip replacement surgery

引用

IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS 2007年第1期E90D卷 30-39页

作者： Kawasaki, Yasuhiro Ino, Fumihiko Sato, Yoshinobu Tamura, Shinichi Hagihara, Kenichi Osaka Univ Grad Sch Informat Sci & Technol Toyonaka Osaka 5608531 Japan Osaka Univ Grad Sch Med Suita Osaka 5650871 Japan

this paper presents the design and implementation of a hip range of motion (ROM) estimation method that is capable of fine-grained estimation during total hip replacement (thR) surgery. Our method is based on two acceleration strategies: (1) adaptive mesh refinement (AMR) for complexity reduction and (2) parallelization for further acceleration. On the assumption that the hip ROM is a single closed region, the AMR strategy reduces the complexity for N x N x N stance configurations from O(N-3) to O(N-D), where 2 <= D <= 3 and D is a data-dependent value that can be approximated by 2 in most cases. the parallelization strategy employs the master-worker paradigm with multiple task queues, reducing synchronization between processors with load balancing. the experimental results indicate that the implementation on a cluster of 64 PCs completes estimation of 360 x 360 x 180 stance configurations in 20 seconds, playing a key role in selecting and aligning the optimal combination of artificial joint components during thR surgery.

关键词： range of motion estimation adaptive mesh refinement cluster computing medical image processing computer assisted surgery

来源：评论

学校读者我要写书评

暂无评论

Time-Decay Attribution Model Using Half-Life in Exponential Decay for E-Commerce Conversion Analysis

Time-Decay Attribution Model Using Half-Life in Exponential ...

引用

international conference on Signal processing and Information Security (ICSPIS)

作者： Debajyoty Banik Suman Kr. Dey Mahendra Kumar Gourisaria School of CS & AI SR University Dept. of Electrical Engineering NIT Rourkela School of Computer Engineering KIIT Deemed to be University

ISBN: (数字)9798350368673

ISBN: (纸本)9798350368680

With the rapid growth of e-commerce platform users, it has become essential for companies to analyze the customer journey leading to a purchase. Users may interact with or be exposed to multiple devices and touchpoints/channels throughout this process. So now, to identify which of those touchpoints resulted in a positive conversion, the companies use various types of attribution models. By identifying this, the company then allocates its marketing spend accordingly in the future on the chosen channels to further optimize its sales and customized user experience. Over time, various types of attribution models have been developed, each model having its pros and cons. this study proposes a novel Time-Decay Attribution Model using Half-life in Exponential Decay (TD-HLED). We also provide a comparative analysis with six different Models: First-touch, Markov Chain Model, Linear-Touch, Last-touch, and our proposed model, TDHLED. Furthermore, extensive experiments have proved that the model attributions in TD-HLED have more accurately reflected how the users have interacted with the marketing than the existing simplistic rule-based architectures.

关键词： Analytical models Data conversion Information security Companies Predictive models Signal processing User experience Data models Delays Electronic commerce

来源：评论

学校读者我要写书评

暂无评论

Accelerated diffusion-based recommendation algorithm on tripartite graphs with GPU clusters

Accelerated diffusion-based recommendation algorithm on trip...

引用

ACIS international conference on Software Engineering, Artificial Intelligence, Networking, and parallel/Distributed Computing (SNPD)

作者： Mi Li Jie Huang Jingpeng Wang School of Software Engineering Tongji University Shanghai China

Exorbitant computation cost hinders the practical application of recommendation algorithm, especially in time-critical application scenario. Although experiments show that recommendation algorithm based on an integrated diffusion on user-item-tag tripartite graphs can significantly improve accuracy, diversification and novelty of recommendation, it is also very time-consuming. therefore, a parallel solution is frequently needed to improve the performance of the algorithm. this paper explicitly presents the parallel implementation and optimizations of diffusion-based recommendation on weighted tripartite graphs algorithm using Compute Unified Device Architecture (CUDA) and related optimization solutions including shared memory, stream scheduling and GPU cluster optimization. Compared to the algorithm running on a single CPU core, the unoptimized GPU kernel can achieve 153.9 speedup on average with the input dataset consists of 30000 records on GTX 980. With shared memory applied, the time cost on memory access saves about 50% on dataset of 90000 records and with 2 way streams scheduling, the kernel's performance improves about 7% ~ 13%. Based on the optimized kernel, we evaluate the performance of the algorithm with customized socket communication mechanism on GPU clusters. And compared to a single GPU node, we achieve 7.55 speedup on clusters of 9 GPUs when recommending for 8000 users. Besides this, the speedup of GPU clusters is also 26.1 times of the speedup of our CPU clusters of 9 nodes and 1586.28 times of serial algorithm on one CPU core. It proves that GPU technology can dramatically improve the algorithm's performance.

关键词： Graphics processing units Kernel Clustering algorithms Bipartite graph Algorithm design and analysis Instruction sets Arrays

来源：评论

学校读者我要写书评

暂无评论

HOI as Embeddings: Advancements of Model Representation Capability in Human-Object Interaction Detection

HOI as Embeddings: Advancements of Model Representation Capa...

引用

IEEE conference on Multimedia Information processing and Retrieval (MIPR)

作者： Junwen Chen Yingcheng Wang Keiji Yanai Department of Informatics The University of Electro-Communications Tokyo Japan

ISBN: (数字)9798350351422

ISBN: (纸本)9798350351439

In recent years, Human-Object Interaction Detection (HOlD) has attracted increasing attention in the computer vision community and has been greatly advanced by the introduction of transformer-based models. However, the representation capability of the pre-trained object detection model is insufficient for capturing the complex interactions between humans and objects, which limits the performance of HOlD methods. In this paper, we introduce three methods to progressively enhance the representation capability. (1) We propose QAHOI to take advantage of multi-scale feature maps with different spatial scales. (2) We propose PQNet to speed up training convergence with parallel queries. (3) We propose SOV-STG to combine the merits of QAHOI and PQNet and introduce the denoising learning strategy to further improve training convergence and performance. Our proposed method SOV-STG achieves state-of-the-art performance on the HICO-DET dataset with one-third of the training epochs compared to previous SOTA methods.

关键词： Training Computer vision Computational modeling Noise reduction Object detection Information processing Computer architecture Transformers Convergence

来源：评论

学校读者我要写书评

暂无评论

An optimal software-pipelining method for instruction-level parallel processors based on scaled retiming

An optimal software-pipelining method for instruction-level ...

引用

international Symposium on Image and Signal processing and Analysis

作者： F. Fernandez A. Sanchez A. Duarte Dept. Fotónica FI-UPM Campus Montegancedo Madrid Spain ESCET-URJC Campus de Móstoles Madrid Spain

ISBN: (纸本)9539676940

Software pipelining is an instruction-level loop scheduling method for achieving high performance fine-grain parallelism on VLIW (very long instruction word) processors. this paper presents a novel software pipelining method for non-pipelining parallel processors based on integer scaling and retiming transformations. this approach generalises and simplifies the analogous extended retiming model of T.W. O'Neil et al. (see Proc. ISCA 12th Int. Conf. parallel & Distributed Computing Syst., p.292-7, 1999; Proc. of ICASSP'99 Conf., vol.4 p.2001-4, 1999). Matrix techniques are used in order to simplify the corresponding graph transformations. Some general properties taken from algebraic graph theory are applied in order to obtain general scheduling techniques: node and cycle methods. the two-phase scheduling method considered is first defined by means of two standard linear programming problems. We transform the corresponding problems into some variants of the maximum cost-to-time ratio problem and shortest path problem, in order to obtain efficient polynomial time algorithms. An example of software pipelining optimization of a digital correlator is also given.

关键词： Pipeline processing Processor scheduling VLIW Software performance parallel processing Distributed computing Graph theory Linear programming Shortest path problem Polynomials

来源：评论

学校读者我要写书评

暂无评论

Hardware-software co-synthesis of low power real-time distributed embedded systems with dynamically reconfigurable FPGAs

Hardware-software co-synthesis of low power real-time distri...

引用

Asia and South Pacific Design Automation conference

作者： L. Shang N.K. Jha Department of EE Princeton University USA

ISBN: (纸本)0769514413

In this paper, we present a multi-objective hardware-software co-synthesis system for multi-rate, real-time, low power distributed embedded systems consisting of dynamically reconfigurable FPGAs, processors, and other system resources. We use an evolutionary algorithm based framework for automatically determining the quantity and type of different system resources, and then assigning tasks to different processing elements (PEs) and task communications to communication links. For FPGAs, we propose a two-dimensional, multi-rate cyclic scheduling algorithm, which determines task priorities based on real-time constraints and reconfiguration overhead information, and then schedules tasks based on the resource utilization and reconfiguration condition in both space and time. the FPGA scheduler is integrated in a list-based system scheduler. To the best of our knowledge, this is the first multi-objective co-synthesis system, which uses dynamically reconfigurable devices to synthesize a distributed embedded system, to target simultaneous optimization of system price and power. Experimental results indicate that our method can reduce schedule length by an average of 41.0% and reconfiguration power by an average of 46.0% compared to the previous method. It also yields multiple system architectures which trade off system price and power under real-time constraints.

关键词： Real time systems Embedded system Field programmable gate arrays Energy consumption Evolutionary computation Resource management Scheduling algorithm Hardware Reconfigurable architectures Contracts

来源：评论

学校读者我要写书评

暂无评论

Structured Pruning for Multi-Task Deep Neural Networks

Structured Pruning for Multi-Task Deep Neural Networks

引用

IEEE conference on Multimedia Information processing and Retrieval (MIPR)

作者： Siddhant Garg Lijun Zhang Hui Guan Adobe Inc. College of Information & Computer Sciences University of Massachusetts Amherst Amherst Massachusetts

ISBN: (数字)9798350351422

ISBN: (纸本)9798350351439

Although multi-task deep neural network (DNN) models have computation and storage benefits over individual single-task DNN models, they can be further optimized via model compression. Numerous structured pruning methods have already been developed that can readily achieve speedups in single-task models, but the pruning of multi-task networks has not yet been extensively studied. In this work, we investigate the effectiveness of structured pruning on multi-task models. We use an existing single-task filter pruning criterion and also introduce an MTL- based filter pruning criterion for estimating the filter importance scores. We prune the model using an iterative pruning strategy with both pruning methods. We show that, with careful hyper-parameter tuning, architectures obtained from different pruning methods do not have significant differences in their performances across tasks when the number of parameters is similar. We also show that iterative structure pruning may not be the best way to achieve a well-performing pruned model because, at extreme pruning levels, there is a high drop in performance across all tasks. But when the same models are randomly initialized and re- trained, they show better results.

关键词： Computational modeling Artificial neural networks Information processing Computer architecture Multitasking Iterative methods Tuning Periodic structures

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：