检索结果-内蒙古大学图书馆

Genetic Programming for Explainable Manifold Learning

IEEE TRANSACTIONS ON EMERGING TOPICS IN computationAL INTELLIGENCE 2025年

作者： Cravens, Ben Lensen, Andrew Maddigan, Paula Xue, Bing Victoria Univ Wellington Ctr Data Sci & Artificial Intelligence Wellington 6140 New Zealand Victoria Univ Wellington Sch Engn & Comp Sci Wellington 6140 New Zealand

Manifold learning techniques play a pivotal role in machine learning by revealing lower-dimensional embeddings within high-dimensional data, thereby enhancing the efficiency, interpretability, and scalability of data analysis. Despite their utility, current manifold learning methods often lack explicit functional mappings, which are critical for ensuring explainability in regulated and high-stakes applications. This paper introduces Genetic Programming for Explainable Manifold Learning (GP-EMaL), a novel integration of Genetic Programming (GP) and Explainable Artificial Intelligence (XAI). GP-EMaL leverages the inherently interpretable, tree-based structures of GP to generate explicit, functional mappings while directly addressing complexity challenges through innovative penalties for tree size, symmetry, and operator selection. By enabling customisable complexity metrics, GP-EMaL adapts to diverse application needs, achieving high manifold quality and significantly improved explainability. Comprehensive experiments demonstrate that GP-EMaL matches or exceeds the performance of existing approaches, producing simpler and more interpretable models. This work advances the state of explainable manifold learning, paving the way for its adoption in domains such as healthcare, environmental modelling, and financial analysis.

关键词： Complexity theory Manifold learning Measurement Optimization Genetic programming computational modeling Feature extraction High dimensional data evolutionary computation Syntactics genetic programming dimensionality reduction explainable artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Binary well placement optimization using a decomposition-based multi-objective evolutionary algorithm with diversity preservation

引用

computationAL GEOSCIENCES 2023年第5期27卷 765-782页

作者： de Moraes, Matheus Bernardelli Coelho, Guilherme Palermo Santos, Antonio Alberto S. Schiozer, Denis Jose Univ Campinas UNICAMP Sch Technol FT Rua Paschoal Marmo1888 Jd Nova Italia BR-13484332 Limeira SP Brazil Univ Campinas UNICAMP Ctr Energy & Petr Studies CEPETRO Rua Cora Coralina 350Cidade Univ BR-13083896 Campinas SP Brazil

In binary multi-objective well placement optimization, multiple conflicting objective functions must be optimized simultaneously in reservoir simulation models containing discrete decision variables. Although multi-objective algorithms have been developed or adapted to tackle this scenario, such as the derivative-free evolutionary algorithms, these methods are known to generate a high number of duplicated strategies in discrete problems. Duplicated strategies negatively impact the optimization process since they: (i) degrade the efficiency of recombination operators in evolutionary algorithms;(ii) slow the convergence speed as they require more iterations to find a well-distributed set of strategies;and (iii) perform unnecessary re-evaluations of previously seen strategies through reservoir simulation. To perform multi-objective well placement optimization while avoiding duplicated strategies, this paper investigates the application of a newly proposed algorithm named MOEA/D-NFTS, with a modified diversity preservation mechanism that incorporates prior knowledge of the problem, on a multi-objective well placement optimization problem. The proposed methodology is evaluated on the UNISIM-II-D benchmark case, a synthetic carbonate black-oil simulation model in a well placement optimization problem using a binary strategy representation, indicating the presence or absence of a given candidate well position in the final strategy. The objective functions are the maximization of the Net Present Value, the maximization of the Cumulative Oil Production, and the minimization of Cumulative Water Production. The modified MOEA/D-NFTS performance is compared with a baseline algorithm without diversity preservation, and the evidence shows that the MOEA/D-NFTS produces statistically significant superior results, and is suitable for binary multi-objective well placement optimization.

关键词： evolutionary computation Field development Management Model-based optimization Production optimization Reservoir simulation

来源：评论

学校读者我要写书评

暂无评论

evolutionary computation Enabled Controlled Charging for E-Mobility Aggregators

Evolutionary Computation Enabled Controlled Charging for E-M...

引用

IEEE Symposium on computational Intelligence Applications in Smart Grid (CIASG)

作者： Hutterer, Stephan Affenzeller, Michael Auinger, Franz Univ Appl Sci Upper Sch Engn & Environm Sci Wels Australia Univ Appl Sci Upper Austria Sch Informat Commun Media Hagenberg Austria

ISBN: (纸本)9781467360029

Optimal integration of electric vehicles (EVs) into modern power grids plays a promising role in future operation of smart power systems. The role of aggregators as e-mobility service providers is getting investigated steadily in recent times and forms a fruitful ground for control of EN' charging. Within this paper, a policy-based control approach is shown that applies an evolutionary simulation optimization procedure for learning valid charging policies offline, that lead to accurate charging decisions online during operation. This approach provides a trade-off between local and distributed control, since the centrally applied learning procedure ensures satisfaction of the operator's requirements during the learning phase, where final control is applied decentrally after distributing the learned policies to the agents. Since the needed information that the aggregator has to provide to the agents is crucial, further analysis on the achieved control policies concerning their data requirements are conducted.

关键词： battery powered vehicles distributed control evolutionary computation learning systems optimisation secondary cells smart power grids

来源：评论

学校读者我要写书评

暂无评论

Computer Highlights Society Magazines

引用

COMPUTER 2024年第11期57卷 4-6页

作者： [Anonymous]

Summaries of article “Computer Highlights Society Magazines.”

关键词： Scientific Applications Computer Graphics Integrated Circuit Design Privacy Convolutional Neural Network Thermal Resistance Additive Manufacturing User Centered evolutionary computation Cyber Physical Systems Role Of Users Comprehensive Safety IT Professionals

来源：评论

学校读者我要写书评

暂无评论

Determination of Weights for Multiobjective Combinatorial Optimization in Incident Management With an evolutionary Algorithm

引用

IEEE ACCESS 2023年 11卷 138502-138514页

作者： Gachnang, Phillip Ehrenthal, Joachim Telesko, Rainer Hanne, Thomas Univ Appl Sci & Arts Northwestern Switzerland Inst Informat Syst CH-4600 Olten Switzerland Univ Appl Sci & Arts Northwestern Switzerland Inst Informat Syst CH-5210 Brugg Switzerland

Incident management in railway operations includes dealing with complex and multiobjective planning problems with numerous constraints, usually with incomplete information and under time pressure. An incident should be resolved quickly with minor deviations from the original plans and at acceptable costs. The problem formulation usually includes multiple objectives relevant to a railway company and the employees involved in controlling operations. Still, there is little established knowledge and agreement regarding the relative importance of objectives such as expressed by weights. Due to the difficulties in assessing weights in a multiobjective context directly involving decision makers, we elaborate on the autoconfiguration of weighted fitness functions based on nine objectives used in a related Integer Linear Programming (ILP) problem. Our approach proposes an evolutionary algorithm and tests it on real-world railway incident management data. The proposed method outperforms the baseline, where weights are equally distributed. Thus, the algorithm shows the capability to learn weights in future applications based on the priorities of employees solving railway incidents which is not yet possible due to an insufficient availability of real-life data from incident management.

关键词： Rail transportation Optimization Linear programming evolutionary computation Software Pareto optimization Safety Path planning evolutionary algorithm incident management multiobjective optimization railway operations weight assessment

来源：评论

学校读者我要写书评

暂无评论

A data-driven approach to neural architecture search initialization

引用

ANNALS OF MATHEMATICS AND ARTIFICIAL INTELLIGENCE 2023年 1-28页

作者： Traore, Kalifou Rene Camero, Andres Zhu, Xiao Xiang Tech Univ Munich Data Sci Earth Observat Arcisstr 21 D-80333 Munich Bavaria Germany German Aerosp Ctr DLR Remote Sensing Inst Munchener Str 20 D-82234 Wessling Bavaria Germany Helmholtz AI Munich Germany

Algorithmic design in neural architecture search (NAS) has received a lot of attention, aiming to improve performance and reduce computational cost. Despite the great advances made, few authors have proposed to tailor initialization techniques for NAS. However, the literature shows that a good initial set of solutions facilitates finding the optima. Therefore, in this study, we propose a data-driven technique to initialize a population-based NAS algorithm. First, we perform a calibrated clustering analysis of the search space, and second, we extract the centroids and use them to initialize a NAS algorithm. We benchmark our proposed approach against random and Latin hypercube sampling initialization using three population-based algorithms, namely a genetic algorithm, an evolutionary algorithm, and aging evolution, on CIFAR-10. More specifically, we use NAS-Bench-101 to leverage the availability of NAS benchmarks. The results show that compared to random and Latin hypercube sampling, the proposed initialization technique enables achieving significant long-term improvements for two of the search baselines, and sometimes in various search scenarios (various training budget). Besides, we also investigate how an initial population gathered on the tabular benchmark can be used for improving search on another dataset, the So2Sat LCZ-42. Our results show similar improvements on the target dataset, despite a limited training budget. Moreover, we analyse the distributions of solutions obtained and find that that the population provided by the data-driven initialization technique enables retrieving local optima (maxima) of high fitness and similar configurations.

关键词： AutoML Neural architecture search evolutionary computation Search Initialization 68Txx

来源：评论

学校读者我要写书评

暂无评论

Convergence Analysis of the Population Learning Algorithm 16th

Convergence Analysis of the Population Learning Algorithm

引用

16th International KES Conference on Intelligent Decision Technologies

作者： Czarnowski, Ireneusz Gdynia Maritime Univ Morska Str 81-87 Gdynia Poland

ISBN: (纸本)9789819774210;9789819774197;9789819774180

In this paper, we review the population learning algorithm and discuss its convergence. This algorithm was proposed as a tool for solving optimisation problems, and the concept underlying this approach is embedded in social educational processes. A convergence analysis of the algorithm is presented by means of a finite Markov chain analysis and by comparing its behaviour to evolutionary strategies in a process of searching for a global solution in a finite number of stages. The proposed population algorithm is also shown to be an alternative tool for solving different optimisation problems.

关键词： Convergence evolutionary computation Population learning algorithm Population-based optimisation tool

来源：评论

学校读者我要写书评

暂无评论

Cardiac arrhythmia discrimination using evolutionary computation

Cardiac arrhythmia discrimination using evolutionary computa...

引用

Computers in Cardiology (CinC)

作者： Juan Francisco Martín-García Inmaculada Mora-Jiménez Arcadio García-Alberola José Luis Rojo-Álvarez Department of Signal Theory and Communications University Rey Juan Carlos Madrid Spain Arrhythmias Unit University Hospital Virgen de la Arrixaca Murcia Spain Prometeo Electric and Electronic Department Universidad de las Fuerzas Armadas ESPE Ecuador

The use of Implantable Cardioverter Defibrillators (ICD) for cardiac arrhythmia treatment implies a search for efficiency in terms of discrimination quality and computational complexity, given that improved efficiency will automatically turn into more effective therapy and longer battery lifetime. In this work, we applied evolutionary computation to create classifiers capable of discriminating between ventricular and supraventricular tachycardia (VT/SVT) in episodes registered by ICDs. evolutionary computation comprises several paradigms emulating natural mechanisms for solving a problem, all of them characterized by a population of individuals (possible solutions) which evolve generation after generation to provide fitter solutions. Genetic programming was the paradigm chosen here because its solutions, coded as decision trees, can be both computationally simple and clinically interpretable. For the experiments, we considered electrograms (EGM) from episodes registered by ICDs in spontaneous/induced tachycardia, previously classified as VT/SVT by clinical experts from several Spanish healthcare centers. Training data were 38 real-valued samples, arranged as the concatenation of two beat segments: a sinus rhythm template immediately previous to the arrhythmic episode (basal reference), and the arrhythmic episode template. Several low complexity trees provided low error rates and allowed physiological interpretation. The best tree yielded an error rate of 1.8%, with both sensitivity and specificity above 98%. This solution compares two samples from the end of the arrhythmic pulse with another two samples from the sinus rhythm, pointing out to a relevant discrimination role of the lasting EGM.

关键词： Error analysis Rhythm Sociology Statistics evolutionary computation Physiology Genetic programming

来源：评论

学校读者我要写书评

暂无评论

An improved binary particle swarm optimization combing V-shaped and U-shaped transfer function

引用

evolutionary INTELLIGENCE 2023年第5期16卷 1653-1666页

作者： Chen, Yuxiang Liu, Jianhua Zhu, Jian Wang, Zihang Fujian Univ Technol Sch Comp Sci & Math Fuzhou Fujian Peoples R China Fujian Univ Technol Fujian Prov Key Lab Big Data Min & Applicat Fuzhou 350118 Fujian Peoples R China

Feature selection aims to find a best feature subset from all feature sets of a given dataset, which represents the whole feature space to reduce redundancy and improve classification accuracy. The evolutionary computation algorithm is often applied to feature selection, but there exists low efficiency in the search process. With the increase of the number of features, solving the feature selection problem become more and more difficult. Existing evolutionary algorithms have many defects, such as slow convergence speed, low convergence accuracy and easy to fall into local optimum. Therefore, the research of more effective evolutionary algorithms has important theoretical significance and application value. Binary Particle Swarm Optimization (BPSO) is a kind of evolutionary computation algorithm and has a good performance in feature selection problems. It uses transfer function to convert the continuous search space to the binary one. Transfer function plays an important role in BPSO. So this paper proposes an improved BPSO by combining V-shaped and U-shaped transfer function, and introduces a new learning strategy and a local search strategy based on adaptive mutation. The improved BPSO enhances its optimization ability in feature selection problem. The experimental results show that the improved BPSO has better dimension reduction ability and classification performance than other algorithms.

关键词： Feature selection Classification Particle swarm optimization (PSO) Transfer function evolutionary computation

来源：评论

学校读者我要写书评

暂无评论

Video summarization support by Interactive evolutionary computation

Video summarization support by Interactive Evolutionary Comp...

引用

International Symposium on Soft Computing and Intelligent Systems (SCIS)

作者： Hiroshi Nakamura Satoshi Ono Department of Information Science and Biomedical Engineering Kagoshima University Japan

This paper proposes a video summarization support method using Interactive evolutionary computation (IEC). The proposed method allows a user, a producer of summarized video, to summarize the video according to the user's preference. To achieve this, this method evaluates a solution on basis of qualitative (user preference) and quantitative (video features) criteria. Experimental results have shown the effectiveness of the proposed method.

关键词： Motion pictures IEC Histograms Image color analysis Color NIST evolutionary computation

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：