检索结果-内蒙古大学图书馆

Histogram-Based estimation of distribution algorithm with RPCL Clustering in Continuous Domain

Histogram-Based Estimation of Distribution Algorithm with RP...

2010 IEEE International Conference on Intelligent Computing and Intelligent Systems(2010 IEEE 智能计算与智能系统国际会议 ICIS 2010)

作者： Hong WU Wei-ping WANG School of Information System and Management National University of Defense Technology Chang Sha P.R.China

Designing efficient estimation of distribution algorithms for optimizing complex continuous problems is still a challenging task. Nowadays, histogram probabilistic model has become a hot topic in the field of estimation of distribution algorithms because of its intrinsic multimodality that makes it proper to describe the solution distribution of complex and multimodal continuous problems. To make histogram probabilistic model more efficiently explore and exploit the search space, rival penalized competitive learning (RPCL) clustering was brought into the algorithm, so that the algorithm could use the knowledge about distribution of values belong to each span. Experimental results showed that the improved algorithm in this paper can give comparable with or better performance than those improved algorithms.

关键词： estimation of distribution algorithm histogram probabilistic model RPCL clustering global optimum elitist strategy

来源：评论

学校读者我要写书评

暂无评论

A Multi-model estimation of distribution algorithm for Agent Routing Problem in Multi-point Dynamic Task

A Multi-model Estimation of Distribution Algorithm for Agent...

引用

第37届中国控制会议

作者： Sai Lu Bin Xin Lihua Dou Ling Wang School of Automation Beijing Institute of Technology Key Laboratory of Intelligent Control and Decision of Complex Systems Beijing Advanced Innovation Center for Intelligent Robots and Systems Department of Automation Tsinghua University

The agent routing problem in multi-point dynamic task（ARP-MPDT） is a multi-task routing problem of a mobile agent. In this problem, there are multiple tasks to be carried out in different locations. As time goes on, the state of each task will change nonlinearly. The agent must go to the task points in turn to perform the tasks, and the execution time of each task is related to the state of the task point when the agent arrives at the point. ARP-MPDT is a typical NP-hard optimization problem. In this paper, we establish the nonlinear ARP-MPDT model. A multi-model estimation of distribution algorithm（EDA） employing node histogram models（NHM） and edge histogram models（EHM） in probability modeling is used to solve the ARP-MPDT. The selection ratio of NHM and EHM probability models is adjusted adaptively. Finally, performance of the algorithm for solving the ARP-MPDT problem is verified by the computational experiments.

关键词： estimation of distribution algorithm Multi-model Routing Multi-point dynamic task

来源：评论

学校读者我要写书评

暂无评论

Design Optimization of Direct Heuristic Dynamic Programming Based on Hybrid estimation of distribution algorithm

Design Optimization of Direct Heuristic Dynamic Programming ...

引用

2015全国理论计算机科学学术年会

作者： Xiong LUO Mi ZHOU Yixuan LV School of Computer and Communication Engineering University of Science and Technology BeijingBeijing 100083China Beijing Key Laboratory of Knowledge Engineering for Materials Science Beijing 100083China

As an important class of approximate dynamic programming, the direct heuristic dynamic programming (DHDP) is discussed in this *** performs well due to its model-free online learning *** the classical DHDP is implemented with gradient-based adaptation learning algorithm of neural network, in this paper we present a design strategy of DHDP with a novel hybrid estimation of distribution algorithm for online learning and control, and the proposed design optimization method achieves the weight training of neural networks with faster convergence *** proposed approach can be viewed as an improvement for *** simulation is conducted on a practical system plant to test the online learning performance by using our ***, the simulation results show the effectiveness of our approach.

关键词： Approximate Dynamic Programming Direct Heuristic Dynamic Programming estimation of distribution algorithm Online Learning

来源：评论

学校读者我要写书评

暂无评论

A New Fuzzy Classification Method Based estimation of distribution algorithm

A New Fuzzy Classification Method Based Estimation of Distri...

引用

2010 International Conference on Services Science, Management and Engineering (SSME 2010)

作者： GAO Xiao-xia The College of computer science and technology,Civil Aviation University of China Tianjin,China HUO Wei-gang The College of computer science and technology,Civil Aviation University of China Tianjin,China

In order to improve fuzzy classification model's accuracy and interpretability,a fuzzy classification method based on estimation of distribution algorithm was *** first constructs initial fuzzy rule set using Apriori principle in the field of data mining,then builds fuzzy classification model by extracting rule from initial fuzzy rule set automatically through Pittsburgh-style binary coding method and UMDA(Univariate Marginal distribution algorithm) estimation of distribution *** experiment on benchmark datasets show that the proposed approach has better performance than fuzzy classification model based on genetic algorithm

关键词： fuzzy classification estimation of distribution algorithm UMDA fuzzy rule

来源：评论

学校读者我要写书评

暂无评论

Global Optimization Using estimation of distribution algorithm

Global Optimization Using Estimation of Distribution Algorit...

引用

2010 2nd International Conference on Intellectual Technology in Industrial Practice (ITIP2010)

作者： YANG Xiuxia, ZHAO Hongchao, ZHANG Yi Department of Control Engineering Naval Aeronautical and Astronautical University Yantai, China

estimation of distribution algorithms (EDA) is a new stochastic optimization algorithm in the field of evolutionary computation. Aiming at the disadvantages of EDA for multi-peak function optimization falling into local optimization easily and not retaining some excellent models, combining with genetic algorithm, the improved EDA is provided. The crossover and mutation operations are added. To maintain the diversity of population, the chaotic initialization is introduced and the individual diversity is adjusted based on the individual density. The new population is produced according to the probability estimation model and the elitist is reserved. A fast parallel EDA with the capacity of global search is designed. Simulation results show that the algorithm can quickly find the global extreme points.

关键词： estimation of distribution algorithm chaotic sequences individual density global optimization

来源：评论

学校读者我要写书评

暂无评论

An estimation of distribution algorithm for resource-constrained project scheduling problem

An estimation of distribution algorithm for resource-constra...

引用

The 22nd China Control and Decision Conference(2010年中国控制与决策会议)

作者： Chen Fang Ling Wang Ye Xu TNList Department of Automation Tsinghua University Beijing 100084

ISBN: (纸本)9781424451814;9781424451821

An estimation of distribution algorithm (EDA) is proposed to solve resource-constrained project scheduling problem (RCPSP). In the EDA, individual is encoded based on the extended active list, and a probability model of the distribution for each activity in a project and its updating mechanism are proposed. The algorithm determines the initial probability matrix according to an initial set of solutions generated by the regret-based sampling method and priority rule, and decodes the individuals by using serial schedule generation scheme. Meanwhile, a permutation based local search method is incorporated into the algorithm to enhance the exploitation ability so as to further improve the searching quality. Simulation results based on benchmarks and comparisons with some existing algorithms demonstrate the feasibility and effectiveness of our proposed EDA.

关键词： estimation of distribution algorithm Resource-constrained project scheduling Probability model

来源：评论

学校读者我要写书评

暂无评论

An estimation of distribution algorithm based on the SWFZIRS operator for global optimization of continuous function

An estimation of distribution algorithm based on the SWFZIRS...

引用

2014 3rd International Conference on Electrical and Electronics Engineering(ICEEE 2014)

作者： Chengcheng Yu Sheng Zhong Li Zhong Department of Network Engineering Hainan College of Software Technology College of Information Science and Technology Hainan University

In this paper, an estimation of distribution algorithm(EDA) is proposed. SWFZIRS(Sector Window Floating Zoom Immune Random Search) operator has been utilized in fitness calculation. The Gaussian distribution has been utilized in the estimation of distribution. Experiments show that the algorithm can improve increasing the probability of obtaining the global optimal solution. It can almost obtain the global optimal solution in two-dimension space.

关键词： evolution algorithm estimation of distribution algorithm SWFZIRS function optimization

来源：评论

学校读者我要写书评

暂无评论

Niching an estimation-of-distribution algorithm by Hierarchical Gaussian Mixture Learning 17

Niching an Estimation-of-Distribution Algorithm by Hierarchi...

引用

Genetic and Evolutionary Computation Conference (GECCO)

作者： Maree, S. C. Alderliesten, T. Thierens, D. Bosman, P. A. N. Acad Med Ctr Amsterdam Netherlands Univ Utrecht Utrecht Netherlands Ctr Wiskunde & Informat Amsterdam Netherlands

ISBN: (纸本)9781450349208

estimation-of-distribution algorithms (EDAs) have been applied with quite some success when solving real-valued optimization problems, especially in the case of Black Box Optimization (BBO). Generally, the performance of an EDA depends on the match between its driving probability distribution and the landscape of the problem being solved. Because most well-known EDAs, including CMA-ES, NES, and AMaLGaM, use a uni-modal search distribution, they have a high risk of getting trapped in local optima when a problem is multi-modal with a (moderate) number of relatively comparable modes. This risk could potentially be mitigated using niching methods that define multiple regions of interest where separate search distributions govern sub-populations. However, a key question is how to determine a suitable number of niches, especially in BBO. In this paper, we present a novel, adaptive niching approach that determines the niches through hierarchical clustering based on the correlation between the probability densities and fitness values of solutions. We test the performance of a combination of this niching approach with AMaLGaM on both new and well-known niching benchmark problems and find that the new approach properly identifies multiple landscape modes, leading to much better performance on multi-modal problems than with a non-niched, uni-modal EDA.y

关键词： continuous optimization black box estimation of distribution algorithm Gaussian mixture model hierarchical clustering

来源：评论

学校读者我要写书评

暂无评论

Automated Analog Synthesis with an estimation of the distribution algorithm 15th

Automated Analog Synthesis with an Estimation of the Distrib...

引用

15th Mexican International Conference on Artificial Intelligence (MICAI)

作者： Torres, Aurora Dolores Torres, Maria Ponce de Leon, Eunice Univ Autonoma Aguascalientes Av Univ 940Ciudad Univ Aguascalientes 20100 AGS Mexico

ISBN: (纸本)9783319624280;9783319624273

This paper presents a set of evolutionary mechanisms embedded on an estimation of distribution algorithm (MITEDA-AC) that performs the synthesis of an analog low pass filter. Analog circuits are modeled with linked lists in order to represent and evolve both, topology and sizing. The developed representation mechanism ensures that generated circuits be feasible, and in order to reduce the gap between real circuits and those evolvable, the concept of preferred values was included on representation and generation mechanisms. The algorithm interacts with SPICE to performance evaluation of each individual in the population. MITEDA-AC was inspired by the COMIT because like this, it uses bivariate probability distributions to generate the optimal dependency tree, but without local optimizers. Features integrated in the learning mechanism of this evolvable algorithm, were the number of capacitors, resistors and inductors included in each circuit of the population. This paper describes the algorithm and discusses its results.

关键词： Automated analog circuit synthesis estimation of distribution algorithm Spice simulation Mutual information Dependency tree

来源：评论

学校读者我要写书评

暂无评论

Random mask-based estimation of the distribution algorithm for stacked auto-encoder one-step pre-training

引用

COMPUTERS & INDUSTRIAL ENGINEERING 2021年 158卷 107400-107400页

作者： Xu, Qingyang Liu, Anbang Yuan, Xianfeng Song, Yong Zhang, Chengjin Li, Yibin Shandong Univ Sch Mech Elect & Informat Engn Jinan 264209 Shandong Peoples R China Shandong Univ Sch Control Sci & Engn Jinan 250100 Shandong Peoples R China

The deep learning techniques have received great achievements in computer vision, natural language processing, etc. The success of deep neural networks depends on the sufficient training of parameters. The traditional way of neural network training is a gradient-based algorithm, which suffers the disadvantage of gradient disappearing, especially for the deeper neural network. Recently, a heuristic algorithm has been proposed for deeper neural network optimization. In this paper, a random mask and elitism univariate continuous estimation of distribution algorithm based on the Gaussian model is proposed to pre-train staked auto-encoder, and then a Stochastic Gradient Descent (SGD) based fine-tuning process is carried out for local searching. In the improved estimation of the distribution algorithm, two individual update strategies are defined;one group of individuals is generated according to the constructed probabilistic model, and another is updated according to the statistics of advanced individuals that aim to reduce the probability of combination explosion and time consumption according to the mask information. In the simulations, different architectures, different mask ratios and different promising individual ratios are adopted to testify the effectiveness of the improved algorithm. According to simulation results, the estimation of thr distribution algorithm has a steady optimization ability for the shallow and stacked autoencoder by one-step pre-training combining SGD based fine-tuning for the MNIST dataset. The proposed model will achieve a state-of-the-art performance on Fashion-MNIST.

关键词： Deep neural network Stacked auto-encoder estimation of distribution algorithm Random mask One-step pre-training

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：