检索结果-内蒙古大学图书馆

26th IEEE International Workshop on Machine Learning for Signal Processing (MLSP)

作者： Chien, Jen-Tzung Huang, Pei-Wen Natl Chiao Tung Univ Dept Elect & Comp Engn Hsinchu Taiwan

ISBN: (纸本)9781509007462

Deep neural network (DNN) is trained according to a mini-batch optimization based on the stochastic gradient descent algorithm. Such a stochastic learning suffers from instability in parameter updating and may easily trap into local optimum. This study deals with the stability of stochastic learning by reducing the variance of gradients in optimization procedure. We upgrade the optimization from the stochastic dual coordinated ascent (SDCA) to the accelerated SDCA without duality (or dual-free ASDCA). This optimization incorporates the momentum method to accelerate the updating rule where the variance of gradients can be reduced. Using dual-free ASDCA, the optimization of dual function of SDCA in a form of convex loss is implemented by directly optimizing the primal function with respect to pseudo-dual parameters. The non-convex optimization in DNN training can be resolved and accelerated. Experimental results illustrate the reduction of training loss, variance of gradients and word error rate by using the proposed optimization for DNN speech recognition.

关键词： optimization algorithm variance reduction deep neural network speech recognition

来源：评论

学校读者我要写书评

暂无评论

A Selection Process for Genetic algorithm Using Clustering Analysis

引用

algorithmS 2017年第4期10卷 123-123页

作者： Chehouri, Adam Younes, Rafic Khoder, Jihan Perron, Jean Ilinca, Adrian Univ Quebec Chicoutimi 555 Blvd Univ Chicoutimi PQ G7H 2B1 Canada Lebanese Univ Fac Engn Branch 3 Rafic Harriri Campus Beirut Lebanon Univ Versailles St Quentin Yvelines LISV Lab 10-12 Ave Europe F-78104 Velizy Villacoublay France Univ Quebec Rimouski WERL 300 Allee Ursulines Rimouski PQ G5L 3A1 Canada

This article presents a newly proposed selection process for genetic algorithms on a class of unconstrained optimization problems. The k-means genetic algorithm selection process (KGA) is composed of four essential stages: clustering, membership phase, fitness scaling and selection. Inspired from the hypothesis that clustering the population helps to preserve a selection pressure throughout the evolution of the population, a membership probability index is assigned to each individual following the clustering phase. Fitness scaling converts the membership scores in a range suitable for the selection function which selects the parents of the next generation. Two versions of the KGA process are presented: using a fixed number of clusters K (KGA(f)) and via an optimal partitioning K-opt (KGA(o)) determined by two different internal validity indices. The performance of each method is tested on seven benchmark problems.

关键词： genetic algorithm selection process clustering k-means optimization algorithm

来源：评论

学校读者我要写书评

暂无评论

Extended shortest path selection for package routing of complex networks

引用

INTERNATIONAL JOURNAL OF MODERN PHYSICS C 2017年第11期28卷 1750129-1750129页

作者： Ye, Fan Zhang, Lei Wang, Bing-Hong Liu, Lu Zhang, Xing-Yi Anhui Univ Dept Comp Sci Hefei Anhui Peoples R China Univ Sci & Technol China Dept Modern Phys Hefei Anhui Peoples R China Univ Sci & Technol China Nonlinear Sci Ctr Hefei Anhui Peoples R China Univ Derby Coll Engn & Technol Comp & Math Dept Derby England Anhui Univ Dept Comp Sci Hefei Anhui Peoples R China

The routing strategy plays a very important role in complex networks such as Internet system and Peer-to-Peer networks. However, most of the previous work concentrates only on the path selection, e.g. Flooding and Random Walk, or finding the shortest path (SP) and rarely considering the local load information such as SP and Distance Vector Routing. Flow-based Routing mainly considers load balance and still cannot achieve best optimization. Thus, in this paper, we propose a novel dynamic routing strategy on complex network by incorporating the local load information into SP algorithm to enhance the traffic flow routing optimization. It was found that the flow in a network is greatly affected by the waiting time of the network, so we should not consider only choosing optimized path for package transformation but also consider node congestion. As a result, the packages should be transmitted with a global optimized path with smaller congestion and relatively short distance. Analysis work and simulation experiments show that the proposed algorithm can largely enhance the network flow with the maximum throughput within an acceptable calculating time. The detailed analysis of the algorithm will also be provided for explaining the efficiency.

关键词： Routing algorithm traffic flow optimization algorithm complex network

来源：评论

学校读者我要写书评

暂无评论

Adaptive niche quantum-inspired immune clonal algorithm

引用

NATURAL COMPUTING 2016年第2期15卷 297-305页

作者： Liu, Jianyong Wang, Huaixiao Sun, Yangyang Li, Ling PLA Univ Sci & Technol Coll Field Engn Nanjing 210007 Jiangsu Peoples R China PLA Univ Sci & Technol Coll Def Engn Nanjing 210007 Jiangsu Peoples R China

The adaptive niche quantum-inspired immune clonal algorithm (ANQICA) is proposed by combining the quantum coding, immune clone and niche mechanism together to solve the multi-modal function optimization more effectively and make the function converge to as many as possible extreme value points. The quantum coding can better explore the solution space, the niche mechanism ensures the algorithm to converge to multi-extremum and the adaptive mechanism is introduced according to the characteristics of each procedure of the algorithm to improve the effect of the algorithm. Example analysis shows that the ANQICA is better in exploration and convergence. Therefore, the ANQICA can be used to solve the problem of multi-modal function optimization effectively.

关键词： optimization algorithm Quantum-inspired evolutionary algorithm Niche Adaptive Immune clonal algorithm

来源：评论

学校读者我要写书评

暂无评论

Belief rule based expert system for classification problems with new rule activation and weight calculation procedures

引用

INFORMATION SCIENCES 2016年 336卷 75-91页

作者： Chang, Leilei Zhou, ZhiJie You, Yuan Yang, Longhao Zhou, Zhiguo High Tech Inst Xian Xian 710025 Shaanxi Peoples R China Fuzhou Univ Decis Sci Inst Fuzhou 350116 Peoples R China Univ Texas SW Med Ctr Dallas Dept Radiat Oncol Dallas TX 75235 USA

Classification problems are significant because they constitute a meta-model for multiple theoretical and practical applications from a wide range of fields. The belief rule based (BRB) expert system has shown potentials in dealing with both quantitative and qualitative information under uncertainty. In this study, a BRB classifier is proposed to solve the classification problem. However, two challenges must be addressed. First, the size of the BRB classifier must be controlled within a feasible range for better expert involvement. Second, the initial parameters of the BRB classifier must be optimized by learning from the experts' knowledge and/or historic data. Therefore, new rule activation and weight calculation procedures are proposed to downsize the BRB classifier while maintaining the matching degree calculation procedure. Moreover, the optimal algorithm using the evidential reasoning (ER) algorithm as the inference engine and the differential evolution (DE) algorithm as the optimization engine is proposed to identify the fittest parameters, including the referenced values of the antecedent attributes, the weights of the rules and the beliefs of the degrees in the conclusion. Five benchmarks, namely, iris, wine, glass, cancer and pima, are studied to validate the efficiency of the proposed BRB classifier. The result shows that all five benchmarks could be precisely modeled with a limited number of rules. The proposed BRB classifier has also shown superior performance in comparing it with the results in the literature. (C) 2015 Elsevier Inc. All rights reserved.

关键词： Classification problems Belief rule base optimization algorithm

来源：评论

学校读者我要写书评

暂无评论

Quadratically Constrained Least Squares with Aerospace Applications

引用

JOURNAL OF GUIDANCE CONTROL AND DYNAMICS 2016年第3期39卷 487-497页

作者： de Ruiter, Anton H. J. Ryerson Univ Dept Aerosp Engn 350 Victoria St Toronto ON M5B 2K3 Canada

This paper treats the problem of quadratically constrained least squares with positive semidefinite weight matrices. A new method of solution is presented that searches directly over the constraint set, and does not require the determination of Lagrange multipliers. Global convergence of the algorithm is rigorously proven. In addition, a covariance analysis is performed for the constrained optimal solution. Two aerospace applications are presented: 1) quadratically constrained Kalman filtering similar in form to the norm-constrained Kalman filter from the literature-it is shown that the optimal quadratically constrained update is simply an orthogonal projection of the optimal unconstrained update onto the constraint set, and 2) a new quadratically constrained Kalman filter using the covariance expression developed in this paper, yielding a statistically more consistent constrained filter. The new filter is demonstrated numerically with a spacecraft attitude estimation example.

关键词： Unscented Kalman Filter Lagrange Multipliers Spacecraft Attitude Computing Quaternions Earth's Magnetic Field optimization algorithm Argument of Latitude Dirac Delta Function Euler Equations

来源：评论

学校读者我要写书评

暂无评论

Evaluation of a transfer function model using experimental data and numerical analysis: the case of a pyroelectric sensor

引用

JOURNAL OF COMPUTATIONAL ELECTRONICS 2016年第3期15卷 1003-1016页

作者： Sifi, Nejmeddine Jridi, Ahmed Touayar, Oualid Univ Carthage Natl Inst Appl Sci & Technol INSAT Tunis Tunisia INSAT Res Lab Mat Measurements & Applicat BP 676 Tunis 1080 Tunisia

The purpose of this paper is to present an evaluation study of a pyroelectric sensor model implemented in MATLAB (R)/SIMULINK (R) environment. The sensor model consists of some cascaded transfer functions taking into account the material characteristics and geometrical parameters which have been synthesized in the form of thermal and electrical time constants and a global multiplying coefficient. The model was proposed by Odon to serve as an excellent basis for the analysis of the dynamic behavior of a pyroelectric sensor. The study performed is relevant to evaluate the validity of the model by comparing its simulated response to measurement obtained on a pyroelectric sensor prototype. To achieve this evaluation, an optimization algorithm is used to estimate the parameters values of the transfer function model using a succession of tests and adjustments so that the algorithm converges to and reaches the optimal solution giving an acceptable correlation between simulated and measured responses. The semi-experimental evaluation approach has been applied to two prototype sensors with pyroelectric material thicknesses of 9 and 25 mu m, respectively. In the two cases, the estimated values of the parameters were very interesting in terms of uncertainties (between 1.846 and 6.726% at 1 sigma level) and setting evidence for the existence of two global solutions, i.e., two deep minima for each prototype sensor.

关键词： Model evaluation Transfer function model Pyroelectric sensor optimization algorithm Measurement

来源：评论

学校读者我要写书评

暂无评论

Optimizing the Heck-Matsuda Reaction in Flow with a Constraint-Adapted Direct Search algorithm

引用

ORGANIC PROCESS RESEARCH & DEVELOPMENT 2016年第11期20卷 1979-1987页

作者： Cortes-Borda, Daniel Kutonova, Ksenia V. Jamet, Corentin Trusova, Marina E. Zammattio, Francoise Truchet, Charlotte Rodriguez-Zubiri, Mireia Felpin, Francois-Xavier Univ Nantes UFR Sci & Tech CNRS UMR 6241 LINA 2 Rue Houssiniere F-44322 Nantes 3 France Univ Nantes UFR Sci & Tech CNRS UMR 6230 CEISAM 2 Rue Houssiniere F-44322 Nantes 3 France Natl Res Tomsk Polytech Univ Dept Biotechnol & Organ Chem Tomsk 634050 Russia Inst Univ France 1 Rue Descartes F-75231 Paris 05 France

The optimization of a palladium-catalyzed Heck Matsuda reaction using an optimization algorithm is presented. We modified and implemented the Nelder-Mead method in order to perform constrained optimizations in a multidimensional space. We illustrated the power of our modified algorithm through the optimization of a multivariable reaction involving the arylation of a deactivated olefin with an arenediazonium salt. The great flexibility of our optimization method allows to fine-tune experimental conditions according to three different objective functions: maximum yield, highest throughput, and lowest production cost. The beneficial properties of flow reactors associated with the power of intelligent algorithms for the fine-tuning of experimental parameters allowed the reaction to proceed in astonishingly simple conditions unable to promote the coupling through traditional batch chemistry.

关键词： flow chemistry palladium Heck reaction optimization algorithm diazonium salts

来源：评论

学校读者我要写书评

暂无评论

Attractive Sets to Unstable Orbits Using Optimal Feedback Control

引用

JOURNAL OF GUIDANCE CONTROL AND DYNAMICS 2016年第12期39卷 2725-2739页

作者： Bando, Mai Scheeres, Daniel J. Kyushu Univ Dept Aeronaut & Astronaut Nishi Ku 744 Motooka Fukuoka 8190395 Japan Univ Colorado Aerosp Engn Sci Boulder CO 80309 USA

This paper investigates the combination of optimal feedback control with the dynamical structure of the three-body problem. The results provide new insights for the design of continuous low-thrust spacecraft trajectories. Specifically, the attracting set of an equilibrium point or a periodic orbit (represented as a fixed point) under optimal control with quadratic cost is obtained. The analysis reveals the relation between the attractive set and original dynamics. In particular, it is found that the largest dimensions of the set are found along the stable manifold and the least extent is along the left eigenvector of the unstable manifold. The asymptotic behavior of the structure of the attractive set when time tends to infinity is analytically revealed. The results generalize the use of manifolds for transfers to equilibrium points and periodic orbits in astrodynamic problems. The result is theoretical and developed for a linearized system, but it can be extended to nonlinear systems in the future.

关键词： Feedback Control Linear Quadratic Control Problem Astrodynamics Spacecraft Trajectories Nonlinear Systems Linear Time Invariant System Halo Orbit Libration Points Earth optimization algorithm

来源：评论

学校读者我要写书评

暂无评论

Gain-Scheduling Control Design in the Presence of Hidden Coupling Terms

引用

JOURNAL OF GUIDANCE CONTROL AND DYNAMICS 2016年第8期39卷 1871-U254页

作者： Lhachemi, H. Saussie, D. Zhu, G. Ecole Polytech Dept Elect Engn Montreal PQ H3T 1J4 Canada

A novel method to handle the hidden coupling terms that naturally arise in the control of a nonlinear plant when an endogenous variable is used as a scheduling parameter was proposed. The proposed approach includes them in the design process with an a priori selection of the scheduling functions. This procedure is generic enough to be applied to a great variety of nonlinear systems in a systematic manner. The efficiency of the proposed approach has been demonstrated on the control of a pitch-axis missile benchmark problem, which has been solved by a nonsmooth optimization algorithm to yield a robust and self-scheduled controller. Numerical simulations show that the resulting controller can achieve a significant performance improvement while satisfying the robustness requirements when faced with a variety of uncertainties.

关键词： State Space Representation Aerodynamic Coefficients Angle of Attack Nonlinear Systems optimization algorithm Nonlinear Feedback Autopilot Numerical Simulation Frequency Domain Missile Control

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：