检索结果-内蒙古大学图书馆

A Pareto-based Ensemble with Feature and Instance Selection for Learning from multi-Class Imbalanced Datasets

INTERNATIONAL JOURNAL OF NEURAL SYSTEMS 2017年第6期27卷 1750028-1750028页

作者： Fernandez, Alberto Jose Carmona, Cristobal Jose del Jesus, Maria Herrera, Francisco Univ Granada Dept Comp Sci & Artificial Intelligence E-18071 Granada Spain Univ Burgos Dept Civil Engn Burgos 09006 Spain De Montfort Univ Leicester Sch Pharm Leicester LE1 9BH Leics England Univ Jaen Dept Comp Sci Jaen 23071 Spain King Abdulaziz Univ Fac Comp & Informat Technol North Jeddah Jeddah 80200 Saudi Arabia

Imbalanced classification is related to those problems that have an uneven distribution among classes. In addition to the former, when instances are located into the overlapped areas, the correct modeling of the problem becomes harder. Current solutions for both issues are often focused on the binary case study, as multi-class datasets require an additional effort to be addressed. In this research, we overcome these problems by carrying out a combination between feature and instance selections. Feature selection will allow simplifying the overlapping areas easing the generation of rules to distinguish among the classes. Selection of instances from all classes will address the imbalance itself by finding the most appropriate class distribution for the learning task, as well as possibly removing noise and difficult borderline examples. For the sake of obtaining an optimal joint set of features and instances, we embedded the searching for both parameters in a multi-objective evolutionary Algorithm, using the C4.5 decision tree as baseline classifier in this wrapper approach. The multi-objective scheme allows taking a double advantage: the search space becomes broader, and we may provide a set of different solutions in order to build an ensemble of classifiers. This proposal has been contrasted versus several state-of-the-art solutions on imbalanced classification showing excellent results in both binary and multi-class problems.

关键词： Imbalanced classification multi-class overlapping feature selection instance selection multi-objective evolutionary algorithms ensembles

来源：评论

学校读者我要写书评

暂无评论

Cooperative coevolutionary multi-guide particle swarm optimization algorithm for large-scale multi-objective optimization problems

引用

SWARM AND evolutionary COMPUTATION 2023年 78卷

作者： Madani, Amirali Engelbrecht, Andries Ombuki-Berman, Beatrice Brock Univ Dept Comp Sci St Catharines ON Canada Stellenbosch Univ Dept Ind Engn Stellenbosch South Africa Stellenbosch Univ Comp Sci Div Stellenbosch South Africa

Many problems that are encountered in real-life applications consist of two or three conflicting objectives and many decision variables. multi-guide particle swarm optimization (MGPSO) is a novel meta-heuristic for multi -objective optimization based on particle swarm optimization (PSO). MGPSO has been shown to be competitive when compared with other state-of-the-art multi-objective optimization algorithms for low-dimensional (and even many-objective) problems. However, a recent study has shown that MGPSO does not scale well when the number of decision variables is increased. This paper proposes a new scalable MGPSO-based algorithm, termed cooperative coevolutionary multi-guide particle swarm optimization (CCMGPSO), that incorporates ideas from cooperative coevolution (CC). CCMGPSO uses new techniques to spend less computational budget by periodically assigning only one CC-based subswarm to each objective (as opposed to using numerous CC -based subswarms). A detailed empirical study on well-known benchmark problems comparing the CCMGPSO with various state-of-the-art large-scale multi-objective optimization algorithms is done. Results show that the proposed CCMGPSO is highly competitive for high-dimensional problems with reference to the inverted generational distance (IGD) metric.

关键词： Particle swarm optimization multi-objective optimization multi-objective evolutionary algorithms Large-scale multi-objective optimization Cooperative coevolution multi-guide particle swarm optimization

来源：评论

学校读者我要写书评

暂无评论

A genetic algorithm-based framework for mining quantitative association rules without specifying minimum support and minimum confidence

引用

SCIENTIA IRANICA 2020年第3期27卷 1316-1332页

作者： Moslehi, F. Haeri, A. Iran Univ Sci & Technol Sch Ind Engn Tehran Iran

Discovering association rules is a useful and common technique for data mining, in which relations and co-dependencies of datasets are shown. One of the most important challenges of data mining is to discover the rules of continuous numerical datasets. Furthermore, another restriction imposed by algorithms in this area is the need to determine the minimum threshold for the support and confidence criteria. In this paper, a multi-objective algorithm for mining quantitative association rules is proposed. The procedure is based on the genetic algorithm, and there is no need to determine the extent of the threshold for the support. and confidence criteria. By proposing a multi-criteria method, useful and attractive rules and the most suitable numerical intervals are discovered, without the need to discretize numerical values and determine the minimum support threshold and minimum confidence threshold. Different criteria are considered to determine appropriate rules. In this algorithm, selected rules are extracted based on confidence, interestingness, and cosine(2). The results obtained from real-world datasets demonstrate the effectiveness of the proposed approach. The algorithm is used to examine three datasets, and the results show the superior performance of the proposed algorithm compared to similar algorithms. (C) 2020 Sharif University of Technology. All rights reserved.

关键词： Machine learning Data mining Quantitative association rules multi-objective evolutionary algorithms Genetic algorithm

来源：评论

学校读者我要写书评

暂无评论

Feature selection based multivariate time series forecasting: An application to antibiotic resistance outbreaks prediction

引用

ARTIFICIAL INTELLIGENCE IN MEDICINE 2020年 104卷 101818-101818页

作者： Jimenez, Fernando Palma, Jose Sanchez, Gracia Marin, David Francisco Palacios, M. D. Lucia Lopez, M. D. Univ Murcia Fac Comp Sci Artificial Intelligence & Knowledge Engn Grp Murcia Spain Univ Hosp Getafe Madrid Spain

Antimicrobial resistance has become one of the most important health problems and global action plans have been proposed globally. Prevention plays a key role in these actions plan and, in this context, we propose the use of Artificial Intelligence, specifically Time Series Forecasting techniques, for predicting future outbreaks of Methicillin-resistant Staphylococcus aureus (MRSA). Infection incidence forecasting is approached as a Feature Selection based Time Series Forecasting problem using multivariate time series composed of incidence of Staphylococcus aureus Methicillin-sensible and MRSA infections, influenza incidence and total days of therapy of both of Levofloxacin and Oseltamivir antimicrobials. Data were collected from the University Hospital of Getafe (Spain) from January 2009 to January 2018, using months as time granularity. The main contributions of the work are the following: the applications of wrapper feature selection methods where the search strategy is based on multi-objective evolutionary algorithms (MOEA) along with evaluators based on the most powerful state-of-the-art regression algorithms. The performance of the feature selection methods has been measured using the root mean square error (RMSE) and mean absolute error (MAE) performance metrics. A novel multi-criteria decision-making process is proposed in order to select the most satisfactory forecasting model, using the metrics previously mentioned, as well as the slopes of model prediction lines in the 1, 2 and 3 steps-ahead predictions. The multi-criteria decision-making process is applied to the best models resulting from a ranking of databases and regression algorithms obtained through multiple statistical tests. Finally, to the best of our knowledge, this is the first time that a feature selection based multivariate time series methodology is proposed for antibiotic resistance forecasting. Final results show that the best model according to the proposed multi-criteria decision making

关键词： Feature selection multi-objective evolutionary algorithms multivariate time series Antibiotic resistance forecasting multiple criteria decision making

来源：评论

学校读者我要写书评

暂无评论

multi-objective SCHEDULING PROBLEM FOR HYBRID MANUFACTURING SYSTEMS WITH WALKING WORKERS

引用

INTERNATIONAL JOURNAL OF INDUSTRIAL ENGINEERING-THEORY APPLICATIONS AND PRACTICE 2019年第5期26卷 625-650页

作者： Yilmaz, Omer Faruk Durmusoglu, Mehmet Bulent Karadeniz Tech Univ Dept Ind Engn Trabzon Turkey Istanbul Tech Univ Dept Ind Engn Istanbul Turkey

Hybrid manufacturing systems (HMSs) are attracting attention from academia and industry owing to concerns regarding fluctuations in customers' demands. An HMS combines traditional manufacturing cells and a functional area, leading to improved flexibility in meeting customers' demands Although information on HMSs is available in the existing academic literature, scheduling problems with walking workers for such systems are rarely addressed. This study explores a multi-objective scheduling problem in HMS and proposes an optimization model to achieve three objectives (i) minimization of average flow time, (ii) reducing the maximum number of workers, and (iii) minimization of the maximum number of workers changing. The model belongs to a non-deterministic polynomial-time hardness (NP-hard) problem class, and hence, non-dominated sorting algorithm-II (NSGA-II) with a local search procedure is employed. The proposed algorithm and several widely accepted multi-objective evolutionary algorithms are compared for six different cases. A set of evaluation metrics is used to evaluate the effectiveness of the proposed algorithm in finding desirable solutions. The computational results indicate that the proposed algorithm is superior to other metaheuristics. This study contributes to existing academic literature by investigating lot scheduling problem with walking workers in the context of hybrid manufacturing systems and provides guidelines for researchers and industry practitioners.

关键词： hybrid manufacturing system scheduling problem walking workers multi-objective evolutionary algorithms NSGA-II

来源：评论

学校读者我要写书评

暂无评论

Finite Satisfiability of Interval Temporal Logic Formulas with multi-objective Metaheuristics

引用

JOURNAL OF multiPLE-VALUED LOGIC AND SOFT COMPUTING 2017年第2-3期28卷 217-249页

作者： Bresolin, Davide Jimenez, Fernando Sanchez, Gracia Sciavicco, Guido Univ Bologna Dept Comp Sci & Engn Bologna Italy Univ Murcia Dept Informat Engn & Commun Murcia Spain

Interval temporal logics provide a natural framework for reasoning about interval structures over linearly ordered domains. Despite being relevant for a broad spectrum of application domains, ranging from temporal databases to artificial intelligence and verification of reactive systems, interval temporal logics still miss tools capable of efficiently supporting them. We approach the finite satisfiability problem for one of the simplest meaningful interval temporal logic, namely A (also known as Right Propositional Neighborhood Logic) and we propose three different multi-objective evolutionary algorithms to solve it by means of a metaheuristic for multi-objective optimization. The resulting semidecision procedure, although incomplete, turns out to be easier to implement and more scalable with respect to classical complete algorithms.

关键词： Interval logics satisfiability metaheuristics multi-objective evolutionary algorithms

来源：评论

学校读者我要写书评

暂无评论

Integrated and dynamic energy modelling of a regional system: A cost-optimized approach in the deep decarbonisation of the Province of Trento (Italy)

引用

ENERGY 2020年 209卷 118378-118378页

作者： Viesi, Diego Crema, Luigi Mahbub, Md Shahriar Verones, Sara Brunelli, Roberto Baggio, Paolo Fauri, Maurizio Prada, Alessandro Bello, Andrea Nodari, Benedetta Silvestri, Silvia Tomasi, Luca Fdn Bruno Kessler FBK Via Sommar 18 I-38123 Povo TN Italy Ahsanullah Univ Sci & Technol AUST Love Rd 141&142 Dhaka 1208 Bangladesh Agenzia Risorse Idriche & lEnergia APRIE Prov Autonoma Trento Piazza Fiera 3 I-38122 Trento TN Italy Univ Trento UNITN Via Mesiano 77 I-38123 Trento TN Italy Fdn Edmund Mach FEM Via E Mach 1 I-38010 San Michele All Adige TN Italy

Since the Kyoto Protocol (1997), the European Union has fought against climate change adopting European, national and regional policies to decarbonise the economy. Moreover, the Paris Agreement (2015) calls 2050 solutions between -80% and -100% of greenhouse gas emissions compared with 1990. Regions have an important role in curbing CO2 emissions, and tailor-made strategies considering local energy demands, savings potentials and renewables must be elaborated factoring in the social and economic context. An "optimized smart energy system" approach is proposed, considering: (I) integration of electricity, thermal and transport sectors, (II) hourly variability of productions and demands, (III) coupling the EnergyPLAN software, to develop integrated and dynamic scenarios, with a multi-objective evolutionary algorithm, to identify solutions optimized both in terms of CO2 emissions and costs, including decision variables for all the three energy sectors simultaneously. The methodology is tested at the regional scale for the Province of Trento (Italy) analyzing a total of 30,000 scenarios. Compared to the Baseline 2016, it is identified: (I) the strategic role of sector coupling among large hydroelectric production and electrification of thermal and transport demands (heat pumps, electric mobility), (II) slight increases in total annual cost, +14% for a -90% of CO2 emissions in 2050. (C) 2020 Elsevier Ltd. All rights reserved.

关键词： Integrated energy system modelling multi-objective evolutionary algorithms Sector coupling Renewable energy Low-carbon economy

来源：评论

学校读者我要写书评

暂无评论

Hybrid NSGA III/dual simplex approach to generation and transmission maintenance scheduling

引用

INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS 2022年 135卷 107498-107498页

作者： San Martin, Luis Salinas Yang, Jin Liu, Ying Univ Glasgow James Watt Sch Engn Glasgow G12 8QQ Lanark Scotland

The generation and transmission maintenance scheduling (GTMS) problem presents generation (GENCOs) and transmission (TRANSCO) companies scheduling their facilities for maintenance to maximize their profits, while the independent system operator (ISO) pushes for maintenance schedules (MS) that guarantees system reliability and minimizes operation cost. Inherently, GTMS is a high-dimensional, non-linear, non-convex, multi-objective optimization problem that contains conflicting objectives related to different participants in the market. This paper develops a hybrid model to tackle the GTMS problem in a deregulated market environment by combining in a novel way the non-dominated sorting genetic algorithm III (NSGA III) and the Dual-Simplex (DS) techniques. The model manages to minimize the total system operational cost and keep high system adequacy, both aspects of interest for the independent system operator (ISO), while increasing the profits of GENCOs. The approach used matches accepted industry maintenance practices with cutting-edge optimization techniques developed in academia. The model, tested in the IEEE-RTS 24 bus test network, delivers a set of feasible MS solutions that address the conflicting relationships between the GENCOs and the ISO in the market, displays a degree of coordination among generation and transmission MS and their impact on electricity prices. Finally, it allows the ISO to use this set to identify the best using the technique for ordering preferences according to the similarity to an ideal solution (TOPSIS) decision-making tool.

关键词： Generation and transmission maintenance scheduling Deregulated electricity market multi-objective optimization multi-objective evolutionary algorithms Non-dominated Sorting Genetic Algorithm Dual-Simplex optimization technique

来源：评论

学校读者我要写书评

暂无评论

Mutation with Local Searching and Elite Inheritance Mechanism in multi-objective Optimization Algorithm: A Case Study in Software Product Line

引用

INTERNATIONAL JOURNAL OF SOFTWARE ENGINEERING AND KNOWLEDGE ENGINEERING 2019年第9期29卷 1347-1378页

作者： Shi, Kai Yu, Huiqun Fan, Guisheng Guo, Jianmei Chen, Liqiong Yang, Xingguang Sun, Huaiying East China Univ Sci & Technol Dept Comp Sci & Engn Shanghai Peoples R China Shanghai Key Lab Comp Software Evaluating & Testi Shanghai Peoples R China Alilibaba Grp Hangzhou Zhejiang Peoples R China Shanghai Inst Technol Dept Comp Sci & Informat Engn Shanghai Peoples R China

An effective method for addressing the configuration optimization problem (COP) in Software Product Lines (SPLs) is to deploy a multi-objective evolutionary algorithm, for example, the state-of-the-art SATIBEA. In this paper, an improved hybrid algorithm, called SATIBEA-LSSF, is proposed to further improve the algorithm performance of SATIBEA, which is composed of a multi-children generating strategy, an enhanced mutation strategy with local searching and an elite inheritance mechanism. Empirical results on the same case studies demonstrate that our algorithm significantly outperforms the state-of-the-art for four out of five SPLs on a quality Hypervolume indicator and the convergence speed. To verify the effectiveness and robustness of our algorithm, the parameter sensitivity analysis is discussed and three observations are reported in detail.

关键词： Software product lines search-based software engineering multi-objective evolutionary algorithms constraint solving

来源：评论

学校读者我要写书评

暂无评论

Software Effort Estimation as a multiobjective Learning Problem

引用

ACM TRANSACTIONS ON SOFTWARE ENGINEERING AND METHODOLOGY 2013年第4期22卷 35-35页

作者： Minku, Leandro L. Yao, Xin Univ Birmingham Ctr Excellence Res Computat Intelligence & Applic Sch Comp Sci Birmingham B15 2TT W Midlands England

Ensembles of learning machines are promising for software effort estimation (SEE), but need to be tailored for this task to have their potential exploited. A key issue when creating ensembles is to produce diverse and accurate base models. Depending on how differently different performance measures behave for SEE, they could be used as a natural way of creating SEE ensembles. We propose to view SEE model creation as a multiobjective learning problem. A multiobjective evolutionary algorithm (MOEA) is used to better understand the tradeoff among different performance-measures by creating SEE models through the simultaneous optimisation of these measures. We show that the performance measures behave very differently, presenting sometimes even opposite trends. They are then used as a source of diversity for creating SEE ensembles. A good tradeoff among different measures can be obtained by using an ensemble of MOEA solutions. This ensemble performs similarly or better than a model that does not consider these measures explicitly. Besides, MOEA is also flexible, allowing emphasis of a particular measure if desired. In conclusion, MOEA can be used to better understand the relationship among performance measures and has shown to be very effective in creating SEE models.

关键词： Experimentation algorithms Management Software effort estimation ensembles of learning machines 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：