检索结果-内蒙古大学图书馆

Incremental learning of bayesian networks with hidden variables

作者： Tian, Fengzhan Zhang, Hongwei Lu, Yuchang Shi, Chunyi State Key Laboratory of Intelligent Technology and System Department of Computer Science and Technology Tsinghua University 100084 Beijing China

ISBN: (纸本)0769511198

In this paper, an incremental method for learning Bayesian networks based on evolutionary computing, IEMA, is put forward. IEMA introduces the evolutionary algorithm and EM algorithm into the process of incremental learning, can not only avoid getting into local maxima, but also incrementally learn Bayesian networks with high accuracy in presence of missing values and hidden variables. In addition, we improved the incremental learning process by Friedman et al. The experimental results verified the validity of IEMA. In terms of storage cost, IEMA is comparable with the incremental learning method of Friedman et al, while it is more accurate. © 2001 IEEE.

关键词： Bayesian networks

来源：评论

学校读者我要写书评

暂无评论

The Suppressed Cavity-Pulling Effect in Dual-Wavelength Active Optical Clock Based on Twice Cavty-Locking Technique

The Suppressed Cavity-Pulling Effect in Dual-Wavelength Acti...

引用

2020 Joint Conference of the IEEE International Frequency Control Symposium and IEEE International Symposium on Applications of Ferroelectrics, IFCS-ISAF 2020

作者： Shi, Tiantian Pan, Duo Chen, Jingbiao State Key Laboratory of Advanced Optical Communication System and Network School of Electronics Engineering and Computer Science Peking University Beijing100871 China

ISBN: (纸本)9781728164304

We have proposed a new scheme for the realization of an excellent active optical frequency standard by utilizing fast and slow twice cavity-locking technique in dual-wavelength active optical clock. The performance is expected to achieve the short term stability of 3.8\times 10^{-16} at 1 s and long-term stability of 10-18 magnitude. Also, the linewidth is expected to reach the quantum limited linewidth with the frequency shift of 7.1\times 10^{-5} \text{Hz}/\mathrm{s}. The system includes four main technologies: dual-wavelength active optical clock, Pound-Drever-Hall technique, phase-stabilized femtosecond laser comb and high speed feedback technique. The cavity length is relocked through slow and fast feedback to two piezoelectric translators of the main-cavity. Therefore, the residual cavity-pulling effect influencing the long-term frequency stability in four-level active optical clock can be suppressed by square times of the bad-cavity coefficient. The frequency stability and the frequency shift of the active optical frequency standard are expected to be two orders and four orders of magnitude better than the ultra-stable reference cavity laser, respectively. © 2020 IEEE.

关键词： Femtosecond lasers

来源：评论

学校读者我要写书评

暂无评论

引用

5th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2001

作者： Ming, Xiao Lu, Yuchang Shi, Chunyi The State Key Laboratory of Intelligent Technology and System Computer Science and Technology Dept Tsinghua University Beijing100084 China

ISBN: (纸本)3540419101

Currently there is no model available that would facilitate the task of finding similar time series based on partial information that interest users. We studied a novel query problem class that we termed micro similarity queries (MSQ) in this paper. We present the formal definition of MSQ. A method is investigated for the purpose of efficient processing of MSQ. We evaluated the behavior of MSQ problem and our query algorithm with both synthetic data and real data. The results show that the knowledge revealed by MSQ corresponds with the subjective feeling of similarity based on singular interest. © Springer-Verlag Berlin Heidelberg 2001.

关键词： Time series analysis

来源：评论

学校读者我要写书评

暂无评论

Tapered photonic crystal fiber for simplified Yb:fiber laser frequency comb with low pulse energy and robust fceo singals

Tapered photonic crystal fiber for simplified Yb:fiber laser...

引用

作者： Jiang, Tongxiao Wang, Aimin Wang, Guizhong Zhang, Wei Niu, Fuzeng Li, Chen Zhang, Zhigang State Key Laboratory of Advanced Optical Communication System and Networks School of Electronics Engineering and Computer Science Peking University Beijing 100871 China

A tapered silica photonic crystal fiber was designed and fabricated to generate more than one octave spanning supercontinuum (from 550 nm to 1400 nm at -30 dB level), by an input pulse of 40 fs 200 pJ directly from an Yb:fiber ring laser. The low pulse energy spectrum broadening are favorable to generate the high contrast fceo signals with low noise. The fceo signal with 40 dB signal-to-noise ratio was detected, which helps to build a compact real-world frequency comb. © 2014 Optical Society of America. © 2014 Optical Society of America.

关键词： Photonic crystal fibers

来源：评论

学校读者我要写书评

暂无评论

Leveraging FVT-margins in design space exploration for FFGA-based CNN accelerators 27

Leveraging FVT-margins in design space exploration for FFGA-...

引用

27th International Conference on Field Programmable Logic and Applications, FPL 2017

作者： Lu, Weina Lu, Wenyan Ye, Jing Hu, Yu Li, Xiaowei State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences China Graduate University of Chinese Academy of Sciences China

ISBN: (纸本)9789090304281

The performance of an FPGA based CNN accelerator is determined by both parallelism and frequency, however, most prior works optimize the parallelism in the RTL design and resolve the frequency after the synthesis. This paper presents a design space exploration method for the pipeline implementation of the deep CNN models, which concurrently optimizes parallelism and frequency to achieve a comprehensive optimization on throughput. In addition to the quantitative modeling on parallelism, the maximum achievable system frequency under various parallelism is explored to leverage the PVT-margins in real-life scenarios and is adopted to guide the design space exploration for further performance boost. A case study of the AlexNet model is implemented using the proposed method on the Altera DE5a-Net board. The experimental results demonstrate that our method can achieve the throughput up to 906.25GOP/s, which gains 1.39× improvement compared to state-of-the-art RTL optimization methods. © 2017 Ghent University.

关键词： Network architecture

来源：评论

学校读者我要写书评

暂无评论

Seamless virtual machine live migration on network security enhanced hypervisor

Seamless virtual machine live migration on network security ...

引用

2009 2nd IEEE International Conference on Broadband Network and Multimedia Technology, IEEE IC-BNMT2009

作者： Xianqin, Chen Han, Wan Sumei, Wang Xiang, Long State Key Laboratory of Virtual Reality Technology System School of Computer Science and Technology Beijing University of Aeronautics and Astronautics Beijing 100191 China

ISBN: (纸本)9781424450053

Since the virtual network traffic is invisible outside the hypervisor, it is impossible for traditional network-base security devices to harness the attacks happened in virtual computing environment. Industry and academies adopt the network security enabled hypervisor (NSE-H) to protect virtual machines (VM) residing in the virtual network. In this paper, we identified the insufficiency of the existing live migration implementation, which prevents itself from providing transparent VM relocation between NSE-Hs. This occurs because the contemporary migration implementation only takes VM encapsulated states into account, but ignores VM related security context(SC) needed by NSE-H embedded security engines (SE). We presented a comprehensive live migration framework for the NSE-H, considering both the execution context encapsulated in VM instance and the VM related security context within the SEs. We built a prototype system of the framework based on stateful firewall enabled Xen hypervisor. Our experiment was performed with realistic applications and the results demonstrate that the solution complements the insufficiency without introducing significant performance downgrade. Even in the worst case, the downtime that occurs during migration increases no more than 15%, comparing to existing implementation. ©2009 IEEE.

关键词： Network security

来源：评论

学校读者我要写书评

暂无评论

LINEAR SYMMETRIC QUANTIZATION OF NEURAL NETWORKS FOR LOW-PRECISION INTEGER HARDWARE 8

LINEAR SYMMETRIC QUANTIZATION OF NEURAL NETWORKS FOR LOW-PRE...

引用

8th International Conference on Learning Representations, ICLR 2020

作者： Zhao, Xiandong Wang, Ying Cai, Xuyi Liu, Cheng Zhang, Lei Institute of Computing Technology Chinese Academy of Sciences China University of Chinese Academy of Sciences China State Key Laboratory of Computer Architecture China

With the proliferation of specialized neural network processors that operate on low-precision integers, the performance of Deep Neural Network inference becomes increasingly dependent on the result of quantization. Despite plenty of prior work on the quantization of weights or activations for neural networks, there is still a wide gap between the software quantizers and the low-precision accelerator implementation, which degrades either the efficiency of networks or that of the hardware for the lack of software and hardware coordination at design-phase. In this paper, we propose a learned linear symmetric quantizer for integer neural network processors, which not only quantizes neural parameters and activations to low-bit integer but also accelerates hardware inference by using batch normalization fusion and low-precision accumulators (e.g., 16-bit) and multipliers (e.g., 4-bit). We use a unified way to quantize weights and activations, and the results outperform many previous approaches for various networks such as AlexNet, ResNet, and lightweight models like MobileNet while keeping friendly to the accelerator architecture. Additional, we also apply the method to object detection models and witness high performance and accuracy in YOLO-v2. Finally, we deploy the quantized models on our specialized integer-arithmetic-only DNN accelerator to show the effectiveness of the proposed quantizer. We show that even with linear symmetric quantization, the results can be better than asymmetric or non-linear methods in 4-bit networks. In evaluation, the proposed quantizer induces less than 0.4% accuracy drop in ResNet18, ResNet34, and AlexNet when quantizing the whole network as required by the integer processors. © 2020 8th International Conference on Learning Representations, ICLR 2020. All rights reserved.

关键词： Chemical activation

来源：评论

学校读者我要写书评

暂无评论

A Handwriting textline extraction approach based on connected domain

A Handwriting textline extraction approach based on connecte...

引用

IEEE International Conference on Cognitive Informatics

作者： Gao, Wei Sun, Fuchun Yin, Zhonghang Department of Computer Science and Technology Tsinghua University Beijing China State Key Laboratory of Intelligent Technology and System Tsinghua University Beijing China

ISBN: (纸本)9781424480401

This paper describes an approach for extracting words, textlines and text blocks by analyzing the spatial configuration of connected domain and word contour rectangles on a given document image. The basic idea is that connected components of black pixels and contours can be used as computational units in document image analysis. In this paper, we try to find a spatial feature and overlapped relationships for every contour rectangle, and we call this feature rectangle "Standard Rectangle"(SR). Then we calculate the split line of every textline according to a series of operations of SRs, and separate the word contour rectangles to different lines. In the next step we estimate that if the adjacent textlines is overlapped. If it is, we calculate the overlap distance and move the word contour rectangles according to it. Our experiment show the approach does good work on both overlapped textlines and detached textlines. © 2010 IEEE.

关键词： Image analysis

来源：评论

学校读者我要写书评

暂无评论

A lightweight hybrid hardware/software approach for object-relative memory profiling

A lightweight hybrid hardware/software approach for object-r...

引用

2012 IEEE International Symposium on Performance Analysis of systems and Software, ISPASS 2012

作者： Chen, Licheng Cui, Zehan Bao, Yungang Chen, Mingyu Huang, Y. Tan, Guangming State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences China Graduate School of Chinese Academy of Sciences China

ISBN: (纸本)9781467311441

Memory profiling is the process of collecting memory address traces during the execution of a program, then analyzing and characterizing the memory behavior of the program offline. With the trend that there will be more and more cores integrated in a processor chip, the Memory Wall problem will become more serious in the chip multiprocessor (CMP) system. Thus accurate and effective memory profiling is becoming one of the keys to identify the source of memory system bottlenecks. A large body of work has been contributed to memory profiling, however, most adopts instrumentation, simulator which suffers heavy overhead, or hardware performance counter which is lack of detail trace information. Furthermore, correlating the raw memory address traces with object-relative information allows us to separate regular pattern for certain object from the irregular mixed, thus helps the optimization. In this paper, we propose a lightweight hybrid hardware/software approach for object-relative memory profiling. We monitor physical memory addresses through hardware snooping with negligible overhead;meanwhile we dump Linux kernel page tables of processes, as well as object-relative memory allocation information. Our approach supports not only to collect applications' full memory traces with detail object relative information, but also to identify hardware-generated memory accesses such as page memory walks due to TLB miss at object level. The experimental results on real system show that our approach is highly accurate (the largest error is 2.04%) and low overhead (the average overhead is 1.60%). Furthermore, we profile two multi-thread applications in detail, and successfully identity hot TLB-miss objects. With object-targeted optimization, we can improve applications' performance by nearly 6.86%. © 2012 IEEE.

关键词： computer operating systems

来源：评论

学校读者我要写书评

暂无评论

Applying system Thinking to Learn from Accident of Modern Automatic Control systems

引用

Chinese Journal of Electronics 2014年第2期23卷 409-414页

作者： NIU Ru CAO Yuan GE Xiaocheng TANG Tao State Key Laboratory of Rail Traffic Control and Safety Beijing Jiaotong University School of Electronics and Information Engineering National Engineering Research Center of Rail Transportation Operation and Control System Beijing Jiaotong University Department of Computer Science University of York

Modern societies increasingly rely on automatic control systems. These systems are hardly pure technical systems; instead they are complex socio-technical systems, which consist of technical elements and social components. It is necessary to have a systematic approach to analyze these systems because it is growing evidence that accidents from these systems usually have complex causal factors which form an interconnected network of events,rather than a simple cause-effect chain. We take railway Train control systems(TCS) as an example to demonstrate the importance of the socio-technical approach to analyze the system. The paper presents an investigation of recent high-speed railway accident by applying STAMP – one of the most notable socio-technical system analysis techniques, outlines improvements to the system which could avoid similar accidents in the future. We also provide our valuable feedback for the use of STAMP.

关键词： Socio-technical automatic system Railway safety Accident analysis systems theoretic accident model and processes(STAMP)

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：