检索结果-内蒙古大学图书馆

4th International Conference on Applied Artificial Intelligence (ICAPAI)

作者： Sorano, Ruslan Ripon, Kazi Shah Nawaz Magnusson, Lars Vidar Halstensen, Thor-David Ghanima, Waleed Ostfold Univ Coll Dept Comp Sci & Commun Halden Norway Oslo Metropolitan Univ Dept Comp Sci Oslo Norway Ostfold Univ Coll Dept Hlth Welf & Org Halden Norway Ostfold Hosp Trust Dept Res Gralum Norway

ISBN: (纸本)9798350349764;9798350349771

Deep vein thrombosis is a serious medical condition requiring prompt and accurate diagnosis. The identification of thrombosis presents a challenging task characterized by conflicting objectives. Maintaining a delicate balance between optimizing overall diagnostic accuracy and averting the misclassification of ill patients as healthy is paramount in the diagnostic process. Our earlier works focused on optimizing the disease prediction accuracy in machine learning models by experimenting with different techniques. We employed single-objective optimization to fine-tune the classification threshold. Additionally, we applied a multi-objective evolutionary algorithm for hyperparameter optimization, both independently and in combination with feature reduction. Expanding on our previous works, this work employs a multi-objective evolutionary algorithm that concurrently tunes hyperparameters, reduces features, and adjusts the classification threshold. By addressing the inherent conflicting objectives in thrombosis diagnostics, the proposed approach generates a set of Pareto-optimal solutions, representing a balance between maximizing overall diagnostic accuracy and minimizing false negatives. Experimental results indicate that this approach enhances the outcomes of the deep vein thrombosis diagnosis prediction, effectively navigating the trade-off in competing objectives for improved clinical efficacy.

关键词： multi-objective evolutionary algorithms healthcare applications decision support deep vein thrombosis hyperparameter tuning feature reduction classification threshold

来源：评论

学校读者我要写书评

暂无评论

Solving multi-objective Optimal Control Problems Using a multiresolution Approach

引用

JOURNAL OF GUIDANCE CONTROL AND DYNAMICS 2025年第1期48卷 32-45页

作者： Parsonage, Ben Maddock, Christie Univ Strathclyde Dept Mech & Aerosp Engn Glasgow Scotland

This paper presents an adaptive multiresolution strategy for multi-objective optimal control problems. The optimal control problem is solved using a direct approach, with individualistic grid adaptation facilitated by a local error analysis at element boundaries. multiple objectives are considered using a dominance-based approach applying both local and global search methods to a collaborative population of unique solutions. These aspects are simultaneously incorporated via a novel application of evolutionary algorithms for adaptive optimal control problems. Together, this avoids the need for a priori specification of the quantity and temporal location of element boundaries and the set of scalarization weights defining the multi-objective descent directions. Solution fidelity can thus increase concurrently with the exploration of the design space, which leads to increased numerical efficiency while propagating and maintaining population diversity. The benefits of the proposed approach over traditional uniform-grid implementations are demonstrated. Results show that the multiresolution approach is capable of striking an effective balance between solution fidelity, population diversity, and computational cost unachievable using uniform grids.

关键词： evolutionary Algorithm Adaptive Mesh Refinement multidisciplinary Design and Optimization Optimization Algorithm Genetic Algorithm multi-objective evolutionary algorithms Optimal Control Problem

来源：评论

学校读者我要写书评

暂无评论

Balancing Performance, Efficiency and Robustness in Open-World Machine Learning via evolutionary multi-objective Model Compression

Balancing Performance, Efficiency and Robustness in Open-Wor...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Del Ser, Javier Martinez-Seras, Aitor Bilbao, Miren Nekane Lobo, Jesus L. Lana, Ibai Herrera, Francisco Basque Res & Technol Alliance BRTA TECNALIA Derio 48160 Bizkaia Spain Univ Basque Country UPV EHU Bilbao 48013 Bizkaia Spain Univ Granada Andalusian Res Inst Data Sci & Computat Intellige Dept Comp Sci & Artificial Intelligence Granada 18071 Spain

ISBN: (纸本)9798350359329;9798350359312

When deploying machine learning models on resource-constrained hardware, reducing the memory footprint required by the model without compromising its performance is critical. Moreover, in open-world scenarios models often operate in dynamic and unpredictable environments where the data distribution evolves over time. Robust models can generalize well to unforeseen circumstances, including out-of-distribution inputs that may not have been encountered during the training phase. This adaptability is essential to handle the inherent variability of real-world data. This work formulates a multi-objective optimization problem that aims at optimizing the quantization resolution of the parameters of an already trained machine learning model based on three conflicting goals: maximizing the performance of the model on its designated learning task, minimizing the memory footprint of the compressed model, and enhancing its robustness against out-of-distribution data. Given the complexity of the resulting combinatorial optimization problem, we employ multi-objective evolutionary algorithms to efficiently obtain an approximation of the Pareto front balancing among the aforementioned objectives. Experiments with a randomized neural network compressed under the proposed formulation are run over several benchmark classification datasets. Different multi-objective solvers are employed to compare their effectiveness in terms of the convergence and diversity of their produced Pareto estimations. Additionally, we assess the achieved equilibrium between the three objectives against a floating-point implementation of the same model. Our experiments reveal that both the computational resources and the robustness of the model can be optimized via evolutionary quantization without significantly sacrificing its performance for the task at hand.

关键词： Model Compression AI Safety Green AI Out-ofDistribution Detection multi-objective evolutionary algorithms Randomization-based Neural Networks.

来源：评论

学校读者我要写书评

暂无评论

Data driven multi-objective optimization of the scheduling for towing a floating offshore wind turbine between assembly port and installation location throughout a year

引用

APPLIED OCEAN RESEARCH 2025年 157卷

作者： Le Pivert, Frederic Roberts, Adam Lopez-Santander, Adan Craven, Matthew J. Kazemi, Saeid Univ Plymouth Sch Engn Comp & Math Maritime Simulat Lab Drake Circus Plymouth PL4 8AA England Univ Plymouth Ctr Math Sci Sch Engn Comp & Math Drake Circus Plymouth PL4 8AA England

High demand for the installation of floating offshore wind turbines over the coming years is likely to place significant pressure on ports and installation vessels. Optimization of the routes between ports and farms and the towing schedule when transporting equipment is therefore critical to reducing operation timescales and carbon emissions. This paper presents two series of multi-objective optimizations for minimizing the timescale and carbon emissions for the case of an IEA 15 MW turbine on a VolturnUS-S platform being wet towed through the English Channel to the Celtic Sea. The study makes use of the Maritime Simulation Laboratory (MSL) Ship Simulator to develop an empirical model of the floating offshore wind turbine being towed under different wind conditions. This is then combined with bathymetry data and historical metocean data from the year 2021 to perform the optimizations. The optimization results are used to feed a second optimization that creates a schedule reducing both emissions and cumulated towing time during a whole year for different number of floating offshore wind turbines.

关键词： Towing Routing optimization multi-objective evolutionary algorithms Floating offshore wind Scheduling

来源：评论

学校读者我要写书评

暂无评论

Subsonic Jet Noise Prediction in Near and Far Field with Optimized Wave-Packet Approach

引用

AIAA JOURNAL 2024年第11期62卷 4144-4152页

作者： Palma, Giorgio Meloni, Stefano Camussi, Roberto Iemma, Umberto Bogey, Christophe CNR Inst Marine Engn CNR INM I-00128 Rome Italy INM Rome Via Vallerano 139 Rome Italy Univ Tuscia Dept Econ Engn Soc & Business Org I-01100 Viterbo Italy Roma Tre Univ Dept Civil Comp Sci & Aeronaut Engn Via Vito Volterra 62 I-00146 Rome Italy Lab Mecan Fluides & Acoust LMFA UMR 5509 F-69130 Ecully France Univ Claude Bernard Lyon I Ecole Cent Lyon CNRS Lyon France

In this study, the parameters of a wave-packet model for subsonic jet noise prediction are systematically optimized by leveraging near- and far-field data obtained from the large-eddy simulation (LES) of a free jet at a Mach number of 0.9 across various radial distances. The utilization of near-field information is justified by the observation that the scattering surfaces are typically situated within a few nozzle diameters from the jet axis in the radial direction, both in the current and in innovative aircraft configurations. The far-field information is used to guarantee the correct subdivision between the wave-packet radiating noise and the hydrodynamic components. The results show a notable agreement between the LES data and the wave-packet solutions, consistent with findings documented in the existing literature. This agreement underscores the validity and applicability of the implemented methodology, offering an effective method for obtaining an equivalent jet noise acoustic source, easily implementable in acoustic scattering codes, and accounting for the directional behavior of jet noise.

关键词： Computational Fluid Dynamics Aircraft Noise Aircraft Components and Structure Jet Noise Aeroacoustics Turbulent-Jet Wave Packets Inverse Problems Acoustic Scattering multi-objective evolutionary algorithms Reduced Order Model

来源：评论

学校读者我要写书评

暂无评论

Development of a multi-objective decision support system for eco-hydrological forest management that quantifies and optimizes ecosystem services related to Carbon, Water, Fire-risk and Eco-resilience (CAFE)

引用

JOURNAL OF ENVIRONMENTAL MANAGEMENT 2025年 380卷 125103页

作者： Perez-Romero, J. Gonzalez-Sanchis, M. Blanco-Cano, L. Del Campo, A. D. Univ Politecn Valencia Res Grp Forest Sci & Technol Re ForeST Cami Vera S-N Valencia 46022 Spain Univ Cordoba Dept Psychol Cordoba 14071 Spain Univ Politecn Valencia Dept Hydraul Engn & Environm Cami Vera S-N Valencia 46022 Spain

Sustainable forest management (FOMA) requires explicit knowledge of the ecosystem services (ES) provided by forests and how they can be improved by different FOMA alternatives, especially when threatened by climate change and the productivity/profitability is low. Decision Support Systems (DSS) have evolved as powerful tools that facilitate decision-making in multi-objective FOMA. However, their use is often limited, as they are monoobjective, are based on empirical relationships, consider limited ES or do not define which silvicultural actions can be deployed, where and when. Thus, decision-makers and forest planners cannot adjust them to local conditions or specific criteria. CAFE (Carbon, Aqua, Fire & Eco-resilience) is a multi-objective DSS for FOMA that quantifies and optimizes different ES that stem from forest management. Its main contribution is coupling ecohydrological process-based models and multi-objective optimization with genetic evolutionary algorithms. The output of the DSS, shown in a user-friendly interface, is a selection of the best-performing solutions (Pareto Optimal Front) of FOMA that optimize the user-selected ES. This tool allows designing and planning silvicultural operations such as thinning or planting required for meeting multiple objectives in FOMA by answering four fundamental questions: How much (thinning intensity or plantation density), where (spatial allocation), when (year of next intervention) and how (target forest strata in which the stand is vertically divided) thinning. This paper presents the design and components of CAFE, and a practical demonstration in contrasted forests. CAFE might contribute to addressing current challenges of meeting global environmental policy goals by stating baselines and additionality in FOMA.

关键词： Process-based models multi-objective evolutionary algorithms Quantification Optimization Forest management

来源：评论

学校读者我要写书评

暂无评论

Knowledge-guided classification and regression surrogates co-assisted multi-objective soft subspace clustering algorithm

引用

APPLIED INTELLIGENCE 2025年第6期55卷 1-29页

作者： Zhao, Feng Li, Lu Liu, Hanqiang Xian Univ Posts & Telecommun Sch Telecommun & Informat Engn Xian 710121 Peoples R China Shaanxi Normal Univ Sch Comp Sci Xian 710119 Peoples R China

The efficiency of multi-objective soft subspace clustering algorithms (MSSCAs) can be low when applied to large-scale datasets. This inefficiency arises because the multi-objective evolutionary algorithms (MOEAs) utilized in MSSCAs often require a large number of soft subspace clustering objective function evaluations due to their population-based nature. Moreover, relying solely on negative Shannon entropy to constrain feature weights is inadequate for soft subspace clustering algorithms. To address these issues, a knowledge-guided classification and regression surrogates co-assisted multi-objective soft subspace clustering (KCRS-MOSSC) algorithm is presented. First, an inter-cluster feature weight dissimilarity function is designed to further constrain the feature weights. Furthermore, a novel surrogate-based optimization framework called the knowledge-guided classification and regression surrogates co-assisted multi-objective evolutionary framework (KCRS-MOEF) is proposed to efficiently optimize the proposed inter-cluster feature weight dissimilarity function, intra-cluster compactness function, inter-cluster separation function, and negative Shannon entropy function. In KCRS-MOEF, a classification decision tree is utilized as the classification surrogate model to help generate a set of promising offspring, while a radial basis function (RBF) model is employed as the regression surrogate model to assist in the infill criterion by predicting the objective function values of the offspring. Furthermore, to fully leverage the knowledge of the evolutionary process, an infill criterion guided by dynamic process knowledge of elite individuals is designed to enhance the convergence and diversity of the population. Finally, a clustering ensemble strategy based on knee point guidance is proposed to generate a final solution from a set of non-dominated individuals. KCRS-MOEF outperforms state-of-the-art counterparts in terms of convergence, diversity, and time efficiency, a

关键词： Subspace clustering multi-objective evolutionary algorithms RBF Surrogate Classification decision tree Surrogate

来源：评论

学校读者我要写书评

暂无评论

Latency-Aware multi-objective Fog Scheduling: Addressing Real-Time Constraints in Distributed Environments

引用

IEEE ACCESS 2024年 12卷 62543-62557页

作者： Altin, Lokman Topcuoglu, Haluk Rahmi Gurgen, Fikret Sadik Bogazici Univ Dept Comp Engn TR-34450 Istanbul Turkiye Siemens Advanta Turkey TR-34870 Istanbul Turkiye Marmara Univ Fac Engn Comp Engn Dept TR-34854 Istanbul Turkiye

The fog computing paradigm was introduced to overcome challenges that cannot be addressed by conventional cloud computing, such as the lower response latency for real-time applications. Task scheduling in fog environments sets forth more complexity using novel objectives beyond scheduling in the cloud. In this study, a task scheduling model with five common objectives and two latency metrics is presented. We propose a latency aware multi-objective multi-rank scheduling algorithm, LAMOMRank, for fog computing. The performance of our algorithm was compared with that of three well known multi-objective scheduling algorithms, Non-dominated Sorting Genetic Algorithm (NSGA-II), Strength Pareto evolutionary Algorithm (SPEA2) and multi-objective Heterogeneous Earliest Finish Time (MOHEFT) algorithm, using three multi-objective metrics and two latency addressing metrics. We populate workload sets using Pegasus workflows and the DeFog benchmark to be distributed over two fog clusters generated with various Amazon Web Services instances. The empirical results validate the significance of our algorithm for better latency fronts including the response latency and task delivery time without performance degradation on multi-objective metrics.

关键词： Fog computing task scheduling latency-constrained applications multi-objective optimization multi-objective evolutionary algorithms directed acyclic graphs

来源：评论

学校读者我要写书评

暂无评论

MoAR-CNN: multi-objective Adversarially Robust Convolutional Neural Network for SAR Image Classification

引用

IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE 2025年第1期9卷 57-74页

作者： Wei, Hai-Nan Zeng, Guo-Qiang Lu, Kang-Di Geng, Guang-Gang Weng, Jian Jinan Univ Coll Cyber Secur Guangzhou 510632 Peoples R China Jinan Univ Natl Joint Engn Res Ctr Network Secur Detect & Pro Guangzhou 510632 Peoples R China Wenzhou Univ Natl Local Joint Engn Res Ctr Digitalized Elect De Wenzhou 325035 Peoples R China Donghua Univ Coll Informat Sci & Technol Shanghai 201620 Peoples R China Zhejiang Univ Inst Cyber Syst & Control Hangzhou Peoples R China

Deep neural networks (DNNs) have been widely applied to the synthetic aperture radar (SAR) images detection and classification recently while different kinds of adversarial attacks from malicious adversary and the hidden vulnerability of DNNs may lead to serious security threats. The state-of-the-art DNNs-based SAR image detection models are designed manually by only considering the test accuracy performance on clean datasets but neglecting the models' adversarial robustness under various types of adversarial attacks. In order to obtain the best trade-off between the clean accuracy and adversarial robustness in robust convolutional neural networks (CNNs)-based SAR image classification models, this work makes the first attempt to develop a multi-objective adversarially robust CNN, called MoAR-CNN. In the MoAR-CNN, we propose a multi-objective automatic design method of the cells-based neural architectures and some critical hyperparameters such as the optimizer type and learning rate of CNNs. A Squeeze-and-Excitation (SE) layer is introduced after each cell to improve the computational efficiency and robustness. The experiments on FUSAR-Ship and OpenSARShip datasets against seven types of adversarial attacks have been implemented to demonstrate the superiority of the proposed MoAR-CNN to six classical manually designed CNNs and four robust neural architectures search methods in terms of clean accuracy, adversarial accuracy, and model size. Furthermore, we also demonstrate the advantages of using SE layer in MoAR-CNN, the transferability of MoAR-CNN, search costs, adversarial training, and the developed NSGA-II in MoAR-CNN through experiments.

关键词： Synthetic aperture radar (SAR) adversarial robustness convolutional neural networks multi-objective evolutionary algorithms hyperparameters optimization neural architectures search Synthetic aperture radar (SAR) adversarial robustness convolutional neural networks multi-objective evolutionary algorithms hyperparameters optimization neural architectures search

来源：评论

学校读者我要写书评

暂无评论

evolutionary multiobjective multicast Virtual Network Function Placement in NFV-SDN Networks 50

Evolutionary Multiobjective Multicast Virtual Network Functi...

引用

50th Latin American Computing Conference (CLEI)

作者： Canete, Carlos Medina, Cristhian More, Luis G. Colbes, Jose Pinto-Roa, Diego P. Univ Nacl Asuncion Fac Politecn Asuncion Paraguay

ISBN: (纸本)9798331540982;9798331540975

Software-defined networking (SDN) and network functions virtualization (NFV) are promising technologies for demand services that require building flexible multicast transmission mechanisms with requirements for data processing functions at the network nodes. The multicast routing problem in NFV-SDN networks seeks to compute multicast-routing trees and place virtual network functions (VNFs), satisfying the traffic demand with optimal resource use and fair data transmission. Since the problem is computationally complex with conflicting objective functions, this paper approaches multicast routing and VNF placement as a multiobjective optimization problem (MOP), minimizing the total resource cost and the maximum transmission delay variance. In this context, this study develops solutions based on multiobjective evolutionary algorithms (MOEAs). Simulations performed on test instances show that the proposals are promissory by computing efficient and non-dominated solutions when compared to a state-of-the-art mono-objective approach.

关键词： multicast Routing Software-Defined Networks Network Functions Virtualization multi-objective evolutionary algorithms

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

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

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：