检索结果-内蒙古大学图书馆

A hybrid neural combinatorial optimization framework assisted by automated algorithm design

COMPLEX & INTELLIGENT SYSTEMS 2024年第6期10卷 8233页

作者： Ma, Liang Hao, Xingxing Zhou, Wei He, Qianbao Zhang, Ruibang Chen, Li Northwest Univ Xian 710127 Peoples R China

In recent years, the application of Neural Combinatorial Optimization (NCO) techniques in Combinatorial Optimization (CO) has emerged as a popular and promising research direction. Currently, there are mainly two types of NCO, namely, the Constructive Neural Combinatorial Optimization (CNCO) and the Perturbative Neural Combinatorial Optimization (PNCO). The CNCO generally trains an encoder-decoder model via supervised learning to construct solutions from scratch. It exhibits high speed in construction process, however, it lacks the ability for sustained optimization due to the one-shot mapping, which bounds its potential for application. Instead, the PNCO generally trains neural network models via deep reinforcement learning (DRL) to intelligently select appropriate human-designed heuristics to improve existing solutions. It can achieve high-quality solutions but at the cost of high computational demand. To leverage the strengths of both approaches, we propose to hybrid the CNCO and PNCO by designing a hybrid framework comprising two stages, in which the CNCO is the first stage and the PNCO is the second. Specifically, in the first stage, we utilize the attention model to generate preliminary solutions for given CO instances. In the second stage, we employ DRL to intelligently select and combine appropriate algorithmic components from improvement pool, perturbation pool, and prediction pool to continuously optimize the obtained solutions. Experimental results on synthetic and real Capacitated Vehicle Routing Problems (CVRPs) and Traveling Salesman Problems(TSPs) demonstrate the effectiveness of the proposed hybrid framework with the assistance of automated algorithm design.

关键词： Neural combinatorial optimization Transformer Reinforcement learning automated algorithm design

来源：评论

学校读者我要写书评

暂无评论

automated algorithm design using proximal policy optimisation with identified features

引用

EXPERT SYSTEMS WITH APPLICATIONS 2023年第1期216卷

作者： Yi, Wenjie Qu, Rong Jiao, Licheng Univ Nottingham Sch Comp Sci Nottingham England Xidian Univ Sch Artificial Intelligence Xian Peoples R China

automated algorithm design is attracting considerable recent research attention in solving complex com-binatorial optimisation problems, due to that most metaheuristics may be particularly effective at certain problems or certain instances of the same problem but perform poorly at others. Within a general algorithm design framework, this study investigates reinforcement learning on the automated design of metaheuristic algorithms. Two groups of features, namely search-dependent and instance-dependent features, are firstly identified to represent the search space of algorithm design to support effective reinforcement learning on the new task of algorithm design. With these key features, a state-of-the-art reinforcement learning technique, namely proximal policy optimisation, is employed to automatically combine the basic algorithmic components within the general framework to develop effective metaheuristics. Patterns of the best designed algorithm, in particular the utilisation and transition of algorithmic components, are investigated. Experimental results on the capacitated vehicle routing problem with time windows benchmark dataset demonstrate the effectiveness of the identified features in assisting automated algorithm design with the proposed reinforcement learning model.

关键词： automated algorithm design Feature identification Reinforcement learning Search pattern Vehicle routing problem

来源：评论

学校读者我要写书评

暂无评论

Multi-objective Human-robot collaborative batch scheduling in distributed hybrid flowshop via automatic design of local search-reconstruction-feedback algorithm

引用

COMPUTERS & INDUSTRIAL ENGINEERING 2025年 203卷

作者： He, Peng Jiang, Xuchu Wang, Qi Zhang, Biao Liaocheng Univ Sch Comp Sci Liaocheng 252000 Peoples R China Zhongnan Univ Econ & Law Sch Stat & Math Wuhan 430073 Peoples R China

The emergence of distributed production models has spurred extensive research on distributed hybrid flowshop scheduling. Despite advancements in resource allocation for flowshops, most studies overlook human-robot collaboration, which remains essential for complex manufacturing processes in real-world production. Additionally, the rise of multi-variety, small-batch production has driven widespread adoption of batch scheduling. Therefore, this paper introduces a multi-objective distributed hybrid flowshop batch scheduling problem with human-robot collaboration (DHFBSP_HC), aiming to minimize makespan and total energy consumption simultaneously. To address this issue, we propose a local search-reconstruction-feedback (LSRF) algorithm, which consists of four core components: population initialization, local search, reconstruction, and feedback mechanism. Additionally, the algorithm incorporates three configurable strategies, including fitness evaluation approaches, initialization approaches, and objective normalization approaches. These configurable strategies are regarded as categorical parameters, whereas the other parameters are referred to as numerical parameters. To select categorical and numerical parameters that can optimize the results of multi-objective DHFBSP_HC, we introduce the automated algorithm design and use I/F-Race to optimize parameter settings. Through comparisons with several state-of-the-art algorithms, we demonstrate the effectiveness and superiority of the LSRF algorithm in solving the multi-objective DHFBSP_HC.

关键词： Distributed hybrid flowshop scheduling Human-robot collaborative scheduling Multi-objective optimization automated algorithm design

来源：评论

学校读者我要写书评

暂无评论

Multi-objective scheduling for surface mount technology workshop: automatic design of two-layer decomposition-based approach

引用

INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH 2025年

作者： Zhang, Biao Wang, Zhi-xuan Meng, Lei-lei Sang, Hong-yan Jiang, Xu-chu Liaocheng Univ Sch Comp Sci Liaocheng Peoples R China Zhongnan Univ Econ & Law Sch Stat & Math Wuhan Peoples R China

In the realm of cellular manufacturing systems (CMS), the scenarios where cells are organised as flowlines have gained substantial practical prevalence. Our focus centres on the domain of the Surface Mount Technology (SMT) workshop, a classical CMS, where cells are harmoniously coordinated to handle intricate production tasks. Two cell-based objectives with a trade-off relationship, namely the number of enabled cells and the makespan among the enabled cells, are introduced. The resulting scheduling problem is referred to as multi-objective reconfigurable distributed flowshop group scheduling problem (MORDFGSP). To tackle this problem, a multi-objective mixed integer programming model is proposed as an analytical tool. Recognising the NP-hard nature of the problem, we develop a two-layer decomposition-based approach that integrates the decomposition-based constructive and improvement heuristics. These heuristics can be configured adopting optional operators for various algorithm components. Within the framework of the developed approach, the automated algorithm design (AAD) is employed to conceive an automated multi-objective algorithm (AMOA) with minimal manual intervention. In the experimental study, the effectiveness of various algorithm components within the approach is thoroughly verified. Furthermore, comparative analyses with alternative methodologies provide strong evidence of the significant superiority of the AMOA.

关键词： Distributed flowshop scheduling group scheduling heuristic multi-objective optimisation automated algorithm design

来源：评论

学校读者我要写书评

暂无评论

Integrated model and automatically designed solver for power system restoration

引用

APPLIED SOFT COMPUTING 2025年 169卷

作者： Zhao, Xiaohui Li, Xia Zhao, Qi Yan, Bai Shi, Yuhui Kang, Jiajin Northeast Elect Power Univ Sch Automat Engn Jilin 132012 Peoples R China Southern Univ Sci & Technol Dept Comp Sci & Engn Shenzhen 518055 Peoples R China Southern Univ Sci & Technol Res Inst Trustworthy Autonomous Syst Shenzhen 518055 Peoples R China

Power system restoration strategies schedule the power plants to re-energize the power network and loads after a blackout. These strategies are of great significance to ensure a resilient power system. Most power system restoration models consider either certain restoration stage(s) or all stages, with each or some of the stages managed independently. They destroy the essential characteristics of the interaction among the restoration stages and lead to sub-optimal solutions. To address this issue, in this paper, we first propose an improved integrated power system restoration (IPSR) model covering the entire restoration process without decoupling and sectionalization. We then develop both mathematical and metaheuristic solvers for the proposed model. In particular, we introduce the automated solver design paradigm to the highly non-linear constrained restoration problem-solving. By the paradigm, we automatically design the metaheuristic solver that gains enhanced performance beyond handcrafted solvers without algorithmic expertise. It is suggested that the automatically designed metaheuristic solver is a promising new option invariable and complicated power system operation scenarios without algorithmic expertise. Finally, a comprehensive case study demonstrates the proposed model's significance and the proposed solver's efficiency. Numerical tests show that the proposed IPSR model and solver obtained over 20% lower restoration time than current restoration methods, demonstrating IPSR's and the proposed solver's efficiency and effectiveness.

关键词： Power system restoration Black start Metaheuristic automated algorithm design

来源：评论

学校读者我要写书评

暂无评论

Evolutionary approach for automated component-based decision tree algorithm design

引用

INTELLIGENT DATA ANALYSIS 2014年第1期18卷 63-77页

作者： Jovanovic, Milos Delibasic, Boris Vukicevic, Milija Suknovic, Milija Martic, Milan Univ Belgrade Fac Org Sci Belgrade Serbia

This paper proposes a framework for automated design of component-based decision tree algorithms. These algorithms are being constructed by interchanging components extracted from decision tree algorithms and their partial improvements. Manual selection of the best-suited algorithm for a specific problem is a complex task because of the huge algorithmic space derived from component-based design. The proposed framework searches through the algorithmic space with an evolutionary algorithm by interchanging components and tuning parameters, and finds a near optimal algorithm for a specific problem. Through experiments we show that using this meta-heuristic is justified in automated component-based algorithm design. This approach is useful not only as an algorithm design help, but also as a technology enhanced learning tool, which aids the understanding of the algorithms.

关键词： Decision tree classification evolutionary algorithm automated algorithm design reusable components component-based algorithms

来源：评论

学校读者我要写书评

暂无评论

AutoOpt: A General Framework for Automatically designing Metaheuristic Optimization algorithms With Diverse Structures

引用

IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE 2025年

作者： Zhao, Qi Yan, Bai Hu, Taiwei Chen, Xianglong Yang, Jian Cheng, Shi Shi, Yuhui Southern Univ Sci & Technol Dept Comp Sci & Engn Shenzhen 518055 Peoples R China Dongguan Univ Technol Sch Comp Sci & Technol Dongguan 523808 Peoples R China Johns Hopkins Univ Dept Comp Sci Baltimore MD 21218 USA Dyearn Technol Co Ltd Shenzhen 518066 Peoples R China Shaanxi Normal Univ Sch Comp Sci Xian 710119 Peoples R China

Metaheuristics are widely recognized gradient-free solvers to hard problems that do not meet the rigorous mathematical assumptions of conventional solvers. The automated design of metaheuristic algorithms provides an attractive path to relieve manual design effort and gain enhanced performance beyond human-made algorithms. However, the specific algorithm prototype and linear algorithm representation in the current automated design pipeline restrict the design within a fixed algorithm structure, which hinders discovering novelties and diversity across the metaheuristic family. To address this challenge, this paper proposes a general framework, AutoOpt, for automatically designing metaheuristic algorithms with diverse structures. AutoOpt contains three innovations: (i) A general algorithm prototype dedicated to covering the metaheuristic family as widely as possible. It promotes high-quality automated design on different problems by fully discovering potentials and novelties across the family. (ii) A directed acyclic graph algorithm representation to fit the proposed prototype. Its flexibility and evolvability enable discovering various algorithm structures in a single run of design, thus boosting the possibility of finding high-performance algorithms. (iii) A graph representation embedding method offering an alternative compact form of the graph to be manipulated, which ensures AutoOpt's generality. Experiments on numeral functions and real applications validate AutoOpt's efficiency and practicability.

关键词： Prototypes Vectors Pipelines Genetic algorithms Computer science Problem-solving Heuristic algorithms Binary trees Space exploration Metaheuristic optimization automated algorithm design automated machine learning evolutionary algorithm evolutionary algorithm

来源：评论

学校读者我要写书评

暂无评论

automated design of Metaheuristics Using Reinforcement Learning Within a Novel General Search Framework

引用

IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION 2023年第4期27卷 1072-1084页

作者： Yi, Wenjie Qu, Rong Jiao, Licheng Niu, Ben Univ Nottingham Sch Comp Sci Computat Optimisat & Learning Lab Nottingham NG8 1BB England Xidian Univ Sch Artificial Intelligence Key Lab Intelligent Percept & Image Understanding Minist Educ Xian 710071 Peoples R China Shenzhen Univ Coll Management Shenzhen 518060 Peoples R China

Metaheuristic algorithms have been investigated intensively to address highly complex combinatorial optimization problems. However, most metaheuristic algorithms have been designed manually by researchers of different expertise without a consistent framework. This article proposes a general search framework (GSF) to formulate in a unified way a range of different metaheuristics. With generic algorithmic components, including selection heuristics and evolution operators, the unified GSF aims to serve as the basis of analyzing algorithmic components for automated algorithm design. With the established new GSF, two reinforcement learning (RL)-based methods, deep Q-network based and proximal policy optimization-based methods, have been developed to automatically design a new general population-based algorithm. The proposed RL-based methods are able to intelligently select and combine appropriate algorithmic components during different stages of the optimization process. The effectiveness and generalization of the proposed RL-based methods are validated comprehensively across different benchmark instances of the capacitated vehicle routing problem with time windows. This study contributes to making a key step toward automated algorithm design with a general framework supporting fundamental analysis by effective machine learning.

关键词： automated algorithm design general search framework metaheuristic reinforcement learning vehicle routing problem

来源：评论

学校读者我要写书评

暂无评论

Sequential Rule Mining for automated design of Meta-heuristics

Sequential Rule Mining for Automated Design of Meta-heuristi...

引用

Genetic and Evolutionary Computation Conference (GECCO)

作者： Meng, Weiyao Qu, Rong Univ Nottingham Sch Comp Sci Nottingham England

ISBN: (纸本)9798400701207

With a recently defined AutoGCOP framework, the design of local search algorithms can be defined as the composition of the basic elementary algorithmic components. These compositions into the best algorithms thus retain useful knowledge of effective algorithm design. This paper investigates effective algorithmic compositions with sequential rule mining techniques to discover valuable knowledge in algorithm design. With the collected effective algorithmic compositions, sequential rules of basic algorithmic components are extracted and further analysed to automatically compose basic algorithmic components within the general AutoGCOP framework to develop new effective meta-heuristics. The sequential rules present superior performance in composing the basic algorithmic components for solving the benchmark vehicle routing problems with time window constraints, demonstrating its effectiveness in designing new algorithms automatically.

关键词： Data Mining Sequential Rule Mining automated algorithm design Meta-heuristics Vehicle Routing Problems

来源：评论

学校读者我要写书评

暂无评论

How Much Data Is Sufficient to Learn High-Performing algorithms? Generalization Guarantees for Data-Driven algorithm design 2021

How Much Data Is Sufficient to Learn High-Performing Algorit...

引用

53rd Annual ACM SIGACT Symposium on Theory of Computing (STOC)

作者： Balcan, Maria-Florina DeBlasio, Dan Dick, Travis Kingsford, Carl Sandholm, Tuomas Vitercik, Ellen Carnegie Mellon Univ Pittsburgh PA 15213 USA Univ Texas El Paso El Paso TX 79968 USA Univ Penn Philadelphia PA 19104 USA Ocean Genom Inc Pittsburgh PA USA Strateg Machine Inc Pittsburgh PA USA Strategy Robot Inc Pittsburgh PA USA Optimized Markets Inc Pittsburgh PA USA

ISBN: (纸本)9781450380539

algorithms often have tunable parameters that impact performance metrics such as runtime and solution quality. For many algorithms used in practice, no parameter settings admit meaningful worst-case bounds, so the parameters are made available for the user to tune. Alternatively, parameters may be tuned implicitly within the proof of a worst-case guarantee. Worst-case instances, however, may be rare or nonexistent in practice. A growing body of research has demonstrated that data-driven algorithm design can lead to significant improvements in performance. This approach uses a training set of problem instances sampled from an unknown, application-specific distribution and returns a parameter setting with strong average performance on the training set. We provide a broadly applicable theory for deriving generalization guarantees that bound the difference between the algorithm's average performance over the training set and its expected performance on the unknown distribution. Our results apply no matter how the parameters are tuned, be it via an automated or manual approach. The challenge is that for many types of algorithms, performance is a volatile function of the parameters: slightly perturbing the parameters can cause a large change in behavior. Prior research (e.g., Gupta and Roughgarden, SICOMP'17;Balcan et al., COLT'17, ICML'18, EC'18) has proved generalization bounds by employing case-by-case analyses of greedy algorithms, clustering algorithms, integer programming algorithms, and selling mechanisms. We uncover a unifying structure which we use to prove extremely general guarantees, yet we recover the bounds from prior research. Our guarantees, which are tight up to logarithmic factors in the worst case, apply whenever an algorithm's performance is a piecewise-constant, -linear, or-more generally-piecewise-structured function of its parameters. Our theory also implies novel bounds for voting mechanisms and dynamic programming algorithms from computational bio

关键词： automated algorithm design data-driven algorithm design automated algorithm configuration machine learning theory computational biology mechanism design

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：