检索结果-内蒙古大学图书馆

Neural network crossover in genetic algorithms using genetic programming

GENETIC PROGRAMMING AND EVOLVABLE MACHINES 2024年第1期25卷 7-7页

作者： Pretorius, Kyle Pillay, Nelishia Univ Pretoria Dept Comp Sci Pretoria South Africa

The use of genetic algorithms (GAs) to evolve neural network (NN) weights has risen in popularity in recent years, particularly when used together with gradient descent as a mutation operator. However, crossover operators are often omitted from such GAs as they are seen as being highly destructive and detrimental to the performance of the GA. Designing crossover operators that can effectively be applied to NNs has been an active area of research with success limited to specific problem domains. The focus of this study is to use genetic programming (GP) to automatically evolve crossover operators that can be applied to NN weights and used in GAs. A novel GP is proposed and used to evolve both reusable and disposable crossover operators to compare their efficiency. Experiments are conducted to compare the performance of GAs using no crossover operator or a commonly used human designed crossover operator to GAs using GP evolved crossover operators. Results from experiments conducted show that using GP to evolve disposable crossover operators leads to highly effectively crossover operators that significantly improve the results obtained from the GA.

关键词： Genetic programming Genetic algorithms Neural networks evolutionary algorithms Crossover operator

来源：评论

学校读者我要写书评

暂无评论

The IGD-based prediction strategy for dynamic multi-objective optimization

引用

SWARM AND evolutionary COMPUTATION 2024年 91卷

作者： Hu, Yaru Peng, Jiankang Ou, Junwei Li, Yana Zheng, Jinhua Zou, Juan Jiang, Shouyong Yang, Shengxiang Li, Jun Xiangtan Univ Dept Comp Sci Coll Xiangtan 411105 Peoples R China Xiangtan Univ Cyberspace Secur Coll Xiangtan 411105 Peoples R China Cent South Univ Dept Automat Changsha Peoples R China Montfort Univ Ctr Computat Intelligence Sch Comp Sci & Informat Leicester LE1 9BH England Hunan Inst Engn Xiangtan 411105 Peoples R China

In recent years, an increasing number of prediction-based strategies have shown promising results in handling dynamic multi-objective optimization problems (DMOPs), and prediction models are also considered to be very favorable. Nevertheless, some linear prediction models may not always be effective. In particular, when the motion direction trends of different individuals are not aligned, these models can yield inaccurate prediction results. Inverted generational distance (IGD) is a commonly used metric for evaluating the performance of algorithms. This paper proposes a prediction model based on the IGD metric. Specifically, we assume that the pareto optimal front (POF) of the population at the previous time step is the true POF, and the POF at the current time step is the approximate POF. We cluster the current population with reference to the euclidean distances from uniform points on the true POF to the current POF points, with slight overlap between adjacent clusters, enables a better tradeoff between convergence and diversity in the prediction process. We consider the movement directions of individuals within each cluster separately through different cluster distributions, while balancing the individual movement directions and the overall population movement direction by overlaying cluster coverage areas, thereby helping to avoid the clustered prediction population from getting trapped in local optima. Experimental results and comparisons with other algorithms demonstrate that this strategy exhibits strong competitiveness in handling DMOPs.

关键词： evolutionary algorithms Dynamic multi-objective optimization Prediction strategy Inverted generational distance

来源：评论

学校读者我要写书评

暂无评论

evolutionary neural networks for deep learning: a review

引用

INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS 2022年第10期13卷 3001-3018页

作者： Ma, Yongjie Xie, Yirong Northwest Normal Univ Coll Phys & Elect Engn Lanzhou 730070 Peoples R China

evolutionary neural networks (ENNs) are an adaptive approach that combines the adaptive mechanism of evolutionary algorithms (EAs) with the learning mechanism of Artificial Neural Network (ANNs). In view of the difficulties in design and development of DNNs, ENNs can optimize and supplement deep learning algorithm, and the more powerful neural network systems are hopefully built. Many valuable conclusions and results have been obtained in this field, especially in the construction of automated deep learning systems. This study conducted a systematic review of the literature on ENNs by using the PRISMA protocol. In literature analysis, the basic principles and development background of ENNs are firstly introduced. Secondly, the main research techniques are introduced in terms of connection weights, architecture design and learning rules, and the existing research results are summarized and the advantages and disadvantages of different research methods are analyzed. Then, the key technologies and related research progress of ENNs are summarized. Finally, the applications of ENNs are summarized and the direction of future work is proposed.

关键词： Deep neural networks evolutionary algorithms evolutionary neural networks Deep learning PRISMA review

来源：评论

学校读者我要写书评

暂无评论

Estimation-of-distribution algorithms for multi-valued decision variables

引用

THEORETICAL COMPUTER SCIENCE 2024年 1003卷

作者： Ben Jedidia, Firas Doerr, Benjamin Krejca, Martin S. Ecole Polytech Inst Polytech Paris F-91120 Palaiseau France Ecole Polytech Inst Polytech Paris Lab Informat LIX CNRS 1 Rue Honore Estienne Orves F-91120 Palaiseau France

The majority of research on estimation -of -distribution algorithms (EDAs) concentrates on pseudoBoolean optimization and permutation problems, leaving the domain of EDAs for problems in which the decision variables can take more than two values, but which are not permutation problems, mostly unexplored. To render this domain more accessible, we propose a natural way to extend the known univariate EDAs to this setting. Different from a na & iuml;ve reduction to the binary case, our approach avoids additional constraints. Since understanding genetic drift is crucial for an optimal parameter choice, we extend the known quantitative analysis of genetic drift to EDAs for multi -valued, categorical variables. Roughly speaking, when the variables take r different values, the time for genetic drift to become significant is r times shorter than in the binary case. Consequently, the update strength of the probabilistic model has to be chosen r times lower now. To investigate how desired model updates take place in this framework, we undertake a mathematical runtime analysis on the r -valued L EADING O NES problem. We prove that with the right parameters, the multi -valued UMDA solves this problem efficiently in O ( r ln( r ) 2 n 2 ln( n )) function evaluations. This bound is nearly tight as our lower bound Omega( r ln( r ) n 2 ln( n )) shows. Overall, our work shows that our good understanding of binary EDAs naturally extends to the multi -valued setting, and it gives advice on how to set the main parameters of multi -values EDAs.

关键词： Estimation-of-distribution algorithms Univariate marginal distribution algorithm evolutionary algorithms Genetic drift LeadingOnes benchmark

来源：评论

学校读者我要写书评

暂无评论

MOEA/D vs. NSGA-II: A Comprehensive Comparison for Multi/Many Objective Analog/RF Circuit Optimization through a Generic Benchmark

引用

ACM TRANSACTIONS ON DESIGN AUTOMATION OF ELECTRONIC SYSTEMS 2024年第1期29卷 1-23页

作者： Saglican, Enes Afacan, Engin Gebze Tech Univ TR-41400 Kocaeli Turkiye

Thanks to the enhanced computational capacity of modern computers, even sophisticated analog/radio frequency (RF) circuit sizing problems can be solved via electronic design automation (EDA) tools. Recently, several analog/RF circuit optimization algorithms have been successfully applied to automatize the analog/RF circuit design process. Conventionally, metaheuristic algorithms are widely used in optimization process. Among various nature-inspired algorithms, evolutionary algorithms (EAs) have been more preferred due to their superiorities (robustness, efficiency, accuracy etc.) over the other algorithms. Furthermore, EAs have been diversified and several distinguished analog/RF circuit optimization approaches for single-, multi-, and many-objective problems have been reported in the literature. However, there are conflicting claims on the performance of these algorithms and no objective performance comparison has been revealed yet. In the previous work, only a few case study circuits have been under test to demonstrate the superiority of the utilized algorithm, so a limited comparison has been made for only these specific circuits. The underlying reason is that the literature lacks a generic benchmark for analog/RF circuit sizing problem. To address these issues, we propose a comprehensive comparison of the most popular two evolutionary computation algorithms, namely Non-Sorting Genetic Algorithm-II and Multi-Objective evolutionary Algorithm based Decomposition, in this article. For that purpose, we introduce two ad hoc testbenches for analog and RF circuits including the common building blocks. The comparison has been made at both multi- and many-objective domains and the performances of algorithms have been quantitatively revealed through the well-known Pareto-optimal front quality metrics.

关键词： Analog RF metaheuristics evolutionary algorithms EDA CAD NSGA-II MOEA/D design automation optimization many-objective testbench

来源：评论

学校读者我要写书评

暂无评论

A hound-inspired pre-hybridized genetic approach for router placement in wireless mesh networks

引用

APPLIED SOFT COMPUTING 2024年 166卷

作者： D'Angelo, Gianni Palmieri, Francesco Univ Salerno Dept Comp Sci Fisciano SA Italy

In the last years, wireless mesh networks (WMNs) have gained more and more popularity in many research and industrial applications thanks to their easy implementation, maintenance, and great reliability at a low cost. Nevertheless, for a large number of nodes, the performance of such networks is heavily influenced by the positioning of routers and gateways over the area to be covered. In this paper, we tackle the router placement problem, which is known to be NP-hard, and its approximate solution through a meta-heuristic approach. The proposed solution empowers the benefits offered by a genetic algorithm pre-hybridized with a local search approach inspired by the behavior of hound dogs. The basic idea is to exploit the dogs' capabilities in moving throughout the solution space to effectively explore it by placing themselves in areas that are more favorable for achieving a high-quality approximate solution in a reasonable time. Experimental results on several benchmarking instances and comparisons with the most effective state-of-the-art algorithms have demonstrated the potential of the proposed approach. This is evidenced by very high connectivity and coverage, a low number of generations, and a small GA population required for convergence. This results in low computational effort and significant time savings, which are of paramount importance in IoT and edge scenarios. We remark that, although offering potential, at the current state, our proposal is not able to adapt to areas with obstacles or irregular shapes.

关键词： Routers placement problem Wireless mesh networks evolutionary algorithms Pre-hybridization Local search

来源：评论

学校读者我要写书评

暂无评论

Surrogate-assisted constraint-handling technique for parametric multi-objective optimization

引用

STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION 2024年第9期67卷 167页

作者： Tsai, Ying-Kuan Malak Jr, Richard J. Northwestern Univ Dept Mech Engn Evanston IL 60208 USA Texas A&M Univ Dept Mech Engn College Stn TX 77843 USA

This work describes a new surrogate-assisted constraint-handling technique (CHT) for parametric multi-objective evolutionary algorithms, called Bayesian CHT. Parametric optimization finds optimal solutions as a function of one or more exogenous variables. The solution is a family of Pareto frontiers called the parameterized Pareto frontier (PPF). CHTs from non-parametric multi-objective evolutionary algorithms do not produce a good sampling of solutions on the PPF. The proposed CHT addresses this using Bayesian methods. A Gaussian process classifier serves as an uncertainty-quantified surrogate for problem constraints, enabling active learning and a novel repair mechanism that promotes sampling along the PPF. The new technique is evaluated on a suite of 36 test problems and two engineering cases of structural design. Quantitative results show that the proposed Bayesian CHT outperforms several state-of-the-art algorithms in most cases. From a qualitative perspective, the nondominated solutions are visualized to support the superiority of the new approach, which achieves a better spread of solutions on the PPF. The results of engineering studies also indicate that the new approach is computationally more efficient than the others.

关键词： Multi-objective Parametric optimization Constraint-handling technique evolutionary algorithms Surrogate modeling Bayesian active learning Test suite

来源：评论

学校读者我要写书评

暂无评论

A non-unit repetitive construction project scheduling with uncertainties

引用

AUTOMATION IN CONSTRUCTION 2024年 164卷

作者： Long, Luong Duc Ho Chi Minh City Univ Technol HCMUT Fac Civil Engn Dept Construct Engn & Management 268 Ly Thuong Kiet St Ho Chi Minh City Vietnam Vietnam Natl Univ Ho Chi Minh City Ho Chi Minh City Vietnam

Repetitive construction project scheduling is a crucial aspect of modern construction project management. This study focuses on the scheduling of non-unit repetitive construction projects with non-serial activity groups, multiple crews, flexible or fixed sequences, and under correlated uncertainties. An integrated model has been developed by combining novel algorithms for non-unit repetitive project scheduling and probabilistic scheduling with correlated uncertainties, alongside evolutionary optimization algorithms (e.g., Differential Evolution (DE), Firefly Algorithm (FA), and the DEFA hybrid). The proposed model is capable of generating near-optimal or optimal schedules with minimal project duration under uncertainties and constraints of work continuity, thereby enhancing the reliability and efficiency of scheduling across various examples. This advancement provides project planners with a valuable tool to manage the complexities of repetitive construction project scheduling under uncertainty. Furthermore, the study lays the groundwork for future research in high-performance computing to enhance optimization techniques and broaden the model's application in construction.

关键词： Non-unit repetitive construction projects evolutionary algorithms Multiple crews Non-serial activity groups Probabilistic scheduling Correlated uncertainties

来源：评论

学校读者我要写书评

暂无评论

Quadratic interpolation and a new local search approach to improve particle swarm optimization: Solar photovoltaic parameter estimation

引用

EXPERT SYSTEMS WITH APPLICATIONS 2024年 236卷

作者： Qaraad, Mohammed Amjad, Souad Hussein, Nazar K. Farag, M. A. Mirjalili, Seyedali Elhosseini, Mostafa A. Abdelmalek Essaadi Univ TIMS FS Tetouan Morocco Amran Univ Fac Sci Dept Comp Sci Amran Yemen Tikrit Univ Coll Comp Sci & Math Dept Math Tikrit Iraq Menoufia Univ Fac Engn Dept Basic Engn Sci Shibin Al Kawm Egypt Torrens Univ Australia Ctr Artificial Intelligence Res & Optimisat Brisbane Australia Yonsei Univ Yonsei Frontier Lab Seoul South Korea Taibah Univ Coll Comp Sci & Engn Yanbu 46421 Saudi Arabia Mansoura Univ Fac Engn Comp & Control Syst Engn Dept Mansoura 35516 Egypt Abdelmalek Essaadi Univ Fac Sci Dept Comp Sci Tetouan Morocco

The Particle Swarm Optimization technique (PSO) is widely used in practical applications due to its flexibility and strong optimization performance. However, like other metaheuristic algorithms, PSO has limitations, such as a propensity to become trapped in local minima and an uneven distribution of effort between exploration and exploitation stages. A novel local search technique called QPSOL, based on PSO is the proposed solution to mitigate these issues. QPSOL aims to increase diversity and achieve a closer balance between the exploration and exploitation phases. The QPSOL incorporates a dynamic optimization strategy to enhance the method's efficiency. Unlike the novel local search strategy, which uses a new local search approach (LSA) to break out of local optima, QPSOL employs quadratic interpolation around the optimal search agent to enhance its exploitation capability and solution accuracy. These strategies complement each other and contribute to boosting PSO's convergence efficiency while seeking to balance exploration and exploitation. The proposed method is assessed using the IEEE CEC'2021 test suite, and its efficacy is evaluated against other metaheuristics and cutting-edge algorithms to determine its trustworthiness. The optimal parameters of three PV models are determined using the proposed technique and compared to different well-established algorithms. Systematic comparisons show that QPSOL is competitive with, and often outperforms, commonly used methods in research for predicting model parameters.

关键词： Hybridization Global optimization Metaheuristic Swarm intelligence evolutionary algorithms Optimization Particle swarm algorithm Photovoltaic system

来源：评论

学校读者我要写书评

暂无评论

Offline learning-based competitive swarm optimizer for non-linear fixed-charge transportation problems

引用

SWARM AND evolutionary COMPUTATION 2024年 88卷

作者： Chauhan, Dikshit Shivani Dr BR Ambedkar Natl Inst Technol Dept Math & Comp Jalandhar 144008 India

The fixed -charge transportation (FCT) problem, an extension of the classical transportation problem, holds significance in practical logistics scenarios where fixed charges play a crucial role. Fixed charges categorize it as an NP -hard problem, posing challenges for conventional methods due to their inefficiency, high computational costs, and susceptibility to local optima. This paper introduces an enhanced evolutionary algorithm, offline learning -based competitive swarm optimization (OLCSO), tailored to address the FCT problems efficiently. OLCSO draws inspiration from offline learning, where particles benefit from the experiences of high performers. The algorithm incorporates (i) a ranking -based mechanism where losers learn from peer -to -peer winners based on their ranks, (ii) a two-phase cooperative evolutionary mechanism enhancing searchability and convergence of the algorithm. In the first phase, losers learn from personal best experiences and the shifted center of the population, boosting diversity. In the second phase, losers learn from winners and the center of the population, improving exploitation, (iii) a mutation strategy is embedded to update winner locations in each evolutionary procedure, and (iv) to ensure a feasible solution to FCT problems, the paper also introduces negative and fractional repair mechanisms. OLCSO is applied to solve two sets of FCT problems, linear and non-linear. Experimental results are compared with various heuristic and metaheuristic techniques, commonly used for these problems, using diverse metrics. A comparative analysis of OLCSO's design elements is also conducted. The results demonstrate OLCSO's superiority over existing algorithms.

关键词： evolutionary algorithms Ranking-based mechanism Offline learning Mutation operator Large-scale Transportation

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：