We introduce a filter-based evolutionaryalgorithm (FEA) for constrained optimization. The filter used by an FEA explicitly imposes the concept of dominance on a partially ordered solution set. We show that the algori...
详细信息
We introduce a filter-based evolutionaryalgorithm (FEA) for constrained optimization. The filter used by an FEA explicitly imposes the concept of dominance on a partially ordered solution set. We show that the algorithm is provably robust for both linear and nonlinear problems and constraints. FEAs use a finite pattern of mutation offsets, and our analysis is closely related to recent convergence results for pattern search methods. We discuss how properties of this pattern impact the ability of an FEA to converge to a constrained local optimum.
Similarity matrix has a significant effect on the performance of the spectral clustering, and how to determine the neighborhood in the similarity matrix effectively is one of its main difficulties. In this paper, a &q...
详细信息
Similarity matrix has a significant effect on the performance of the spectral clustering, and how to determine the neighborhood in the similarity matrix effectively is one of its main difficulties. In this paper, a "divide and conquer" strategy is proposed to model the similarity matrix construction task by adopting multiobjective evolutionary algorithm (MOEA). The whole procedure is divided into two phases, phase I aims to determine the nonzero entries of the similarity matrix, and Phase II aims to determine the value of the nonzero entries of the similarity matrix. In phase I, the main contribution is that we model the task as a biobjective dynamic optimization problem, which optimizes the diversity and the similarity at the same time. It makes each individual determine one nonzero entry for each sample, and the encoding length decreases to O(N) in contrast with the non-ensemble multiobjective spectral clustering. In addition, a specific initialization operator and diversity preservation strategy are proposed during this phase. In phase II, three ensemble strategies are designed to determine the value of the nonzero value of the similarity matrix. Furthermore, this Pareto ensemble framework is extended to semi-supervised clustering by transforming the semi-supervised information to constraints. In contrast with the previous multiobjectiveevolutionary-based spectral clustering algorithms, the proposed Pareto ensemble-based framework makes a balance between time cost and the clustering accuracy, which is demonstrated in the experiments section.
The operation process design is one of the key issues in the manufacturing and service sectors. As a typical operation process, the scheduling with consideration of the deteriorating effect has been widely studied;how...
详细信息
The operation process design is one of the key issues in the manufacturing and service sectors. As a typical operation process, the scheduling with consideration of the deteriorating effect has been widely studied;however, the current literature only studied single function requirement and rarely considered the multiple function requirements which are critical for a real-world scheduling process. In this article, two function requirements are involved in the design of a scheduling process with consideration of the deteriorating effect and then formulated into two objectives of a mathematical programming model. A novel multiobjective evolutionary algorithm is proposed to solve this model with combination of three strategies, i.e. a multiple population scheme, a rule-based local search method and an elitist preserve strategy. To validate the proposed model and algorithm, a series of randomly-generated instances are tested and the experimental results indicate that the model is effective and the proposed algorithm can achieve the satisfactory performance which outperforms the other state-of-the-art multiobjective evolutionary algorithms, such as nondominated sorting genetic algorithm II and multiobjective evolutionary algorithm based on decomposition, on all the test instances.
Convolutional neural networks (CNNs) have achieved state-of-the-art performance in many image classification tasks. However, training a deep CNN requires a massive amount of training data, which can be expensive or un...
详细信息
Convolutional neural networks (CNNs) have achieved state-of-the-art performance in many image classification tasks. However, training a deep CNN requires a massive amount of training data, which can be expensive or unobtainable in practical applications, such as defect inspection and medical diagnosis. Transfer learning has been developed to address this issue by transferring knowledge learned from source domains to target domains. A common approach is fine-tuning, which adapts the parameters of a trained neural network for the new target task. Nevertheless, the network architecture remains designed for the source task rather than the target task. To optimize the network architecture in transfer learning, we propose a two-stage evolutionary neural architecture search for transfer learning (EvoNAS-TL), which searches for an efficient subnetwork of the source model for the target task. EvoNAS-TL features two search stages: 1) structure search and 2) local enhancement. The former conducts a coarse-grained global search for suitable neural architectures, while the latter acts as a fine-grained local search to refine the models obtained. In this study, neural architecture search (NAS) is formulated as a multiobjective optimization problem that concurrently minimizes the prediction error and model size. The knee-guided multiobjective evolutionary algorithm, a modern multiobjective optimization approach, is employed to solve the NAS problem. In this study, several experiments are conducted to examine the effectiveness of EvoNAS-TL. The results show that applying EvoNAS-TL on VGG-16 can reduce the model size by 52%-85% and simultaneously improve the testing accuracy by 0.7%-6.9% in transferring from ImageNet to CIFAR-10 and NEU surface detection datasets. In addition, EvoNAS-TL performs comparably to or better than state-of-the-art methods on the CIFAR-10, NEU, and Office-31 datasets.
The existing parallel multiobjectiveevolutionary computation does not perform well for constrained multiobjective optimization problems with discontinuous Pareto fronts or narrow feasible regions. This study parallel...
详细信息
The existing parallel multiobjectiveevolutionary computation does not perform well for constrained multiobjective optimization problems with discontinuous Pareto fronts or narrow feasible regions. This study parallelizes the state-of-the-art cooperative multiobjective coevolutionaryalgorithm and proposes an effective parallel evolutionaryalgorithm for constrained multiobjective optimization problems that are difficult to optimize. Two parallelization methods are compared: a global parallel model in which solution evaluations are performed in parallel, and a hybrid model that treats the cooperative populations in a distributed manner while performing each solution evaluation in parallel. The first model is a straightforward parallelization, while the second one capitalizes on the characteristics of the coevolutionary framework. To investigate the efficacy of the proposed models, experiments are conducted on constrained multiobjective optimization problems, including complex characteristics, while varying the number of parallel cores up to 64. The experiments compare the two proposed methods from the viewpoint of search performance and execution time. The experimental results reveal that the latter hybrid model shows better computational efficiency and scalability against an increasing number of cores without adversely affecting the search performance compared to the former straightforward parallelization.
Most of existing image segmentation algorithms are only based on the color feature. However, the spatial distribution of an image can not be well described by using the color feature alone. Thus, it is necessary to ad...
详细信息
Most of existing image segmentation algorithms are only based on the color feature. However, the spatial distribution of an image can not be well described by using the color feature alone. Thus, it is necessary to add additional features to design efficient segmentation algorithms. Although many researchers also try to use multiple features for image segmentation, it is extremely difficult to combine multiple features automatically. This paper proposes a multiojective multiple features fusion strategy for image segmentation. The basic idea is to convert the segmentation problem into a multiobjective optimization problem, in which each objective considers one feature. It contains three steps. First, the original image is split into a set of over-segmented regions by using Meanshift to preserve the spatial details and to simplify the segmentation problem. Second, both the color and texture features are extracted to describe the regions. And two similarity matrices are designed by computing the similarity between each pair of regions in two features respectively. Third, a multiobjectiveevolutionary clustering algorithm is applied to merge these over-segmented regions. In this stage, two objective functions are designed based on the color and texture features respectively. A region index encoding scheme is introduced to design the individual, which contains some cluster representative regions. Some evolutionary operators are proposed to generate the new population. In the final generation, the best solution is selected from nondominated solutions for subsequent segmentation. Experiment results show that the proposed method provides promising segmentation results in combining the color and texture features.
The RNA inverse folding problem involves discovering a nucleotide sequence that folds into a desired target structure. Although numerous computational methods have been proposed over the years to tackle the problem, n...
详细信息
The RNA inverse folding problem involves discovering a nucleotide sequence that folds into a desired target structure. Although numerous computational methods have been proposed over the years to tackle the problem, none have successfully solved the complete Eterna100 set. The Eterna100 set is widely recognized as a benchmark in this field. Therefore, there is still ample room for improvement in this area. This paper aims to address this challenge by introducing eM2dRNAs, an enhanced version of our previous approach called m2dRNAs, which is a multiobjective metaheuristic to design RNA sequences. By introducing eM2dRNAs, we aim to make significant advancements in RNA inverse folding. Our approach starts with the recursive decomposition of the target structure, simplifying the problem to be solved. We conducted a comparative study of our method against several published methods using the Eterna100 benchmark. The results showed that our proposal performs significantly better than the other methods across almost all metrics and categories considered, thus achieving our objective of improving the ability to solve the RNA inverse folding problem. (c) 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://***/licenses/by-nc-nd/4.0/).
This article compares three multiobjective evolutionary algorithms (MOEAs) with application to the urban drainage system (UDS) adaptation of a capital city in North China. Particularly, we consider the well-known NSGA...
详细信息
This article compares three multiobjective evolutionary algorithms (MOEAs) with application to the urban drainage system (UDS) adaptation of a capital city in North China. Particularly, we consider the well-known NSGA-II, the built-in solver in the MATLAB Global Optimization Toolbox (MLOT), and a newly-developed hybrid MOEA called GALAXY. A variety of parameter combinations of each MOEA is systemically applied to examine their impacts on optimization efficiency. Results suggest that the traditional MOEAs suffer from severe parameterization issues. For NSGA-II, the distribution indexes of crossover and mutation operators were found to have dominant impacts, while the probabilities of the two operators played a secondary role. For MLOT, the two-point and the scattered crossover operators accompanied by the adaptive-feasible mutation operator gained the best Pareto fronts, provided the crossover fraction is set to lower values. In contrast, GALAXY was the most robust and easy-to-use tool among the three MOEAs, owing to its elimination of various associated parameters of searching operators for substantially alleviating the parameterization issues. This study contributes to the literature by showing how to improve the robustness of identifying optimal solutions through better selection of operators and associated parameter settings for real-world UDS applications.
Detecting communities of complex networks has been an effective way to identify substructures that could correspond to important functions. Conventional approaches usually consider community detection as a single-obje...
详细信息
Detecting communities of complex networks has been an effective way to identify substructures that could correspond to important functions. Conventional approaches usually consider community detection as a single-objective optimization problem, which may confine the solution to a particular community structure property. Recently, a new community detection paradigm is emerging: multiobjective optimization for community detection, which means simultaneously optimizing multiple criteria and obtaining a set of community partitions. The new paradigm has shown its advantages. However, an important issue is still open: what type of objectives should be optimized to improve the performance of multiobjective community detection? To exploit this issue, we first proposed a general multiobjective community detection solution (called NSGA-Net) and then analyzed the structural characteristics of communities identified by a variety of objective functions that have been used or can potentially be used for community detection. After that, we exploited correlation relations (i.e., positively correlated, independent, or negatively correlated) between any two objective functions. Extensive experiments on both artificial and real networks demonstrate that NSGA-Net optimizing over a pair of negatively correlated objectives usually leads to better performances compared with the single-objective algorithm optimizing over either of the original objectives, or even to other well-established community detection approaches.
Optimal management policies for water reservoir operation are generally designed via stochastic dynamic programming (SDP). Yet, the adoption of SDP in complex real-world problems is challenged by the three curses of d...
详细信息
Optimal management policies for water reservoir operation are generally designed via stochastic dynamic programming (SDP). Yet, the adoption of SDP in complex real-world problems is challenged by the three curses of dimensionality, modeling, and multiple objectives. These three curses considerably limit SDP's practical application. Alternatively, this study focuses on the use of evolutionarymultiobjective direct policy search (EMODPS), a simulation-based optimization approach that combines direct policy search, nonlinear approximating networks, and multiobjective evolutionary algorithms to design Pareto-approximate closed-loop operating policies for multipurpose water reservoirs. This analysis explores the technical and practical implications of using EMODPS through a careful diagnostic assessment of the effectiveness and reliability of the overall EMODPS solution design as well as of the resulting Pareto-approximate operating policies. The EMODPS approach is evaluated using the multipurpose Hoa Binh water reservoir in Vietnam, where water operators are seeking to balance the conflicting objectives of maximizing hydropower production and minimizing flood risks. A key choice in the EMODPS approach is the selection of alternative formulations for flexibly representing reservoir operating policies. This study distinguishes between the relative performance of two widely-used nonlinear approximating networks, namely artificial neural networks (ANNs) and radial basis functions (RBFs). The results show that RBF solutions are more effective than ANN ones in designing Pareto approximate policies for the Hoa Binh reservoir. Given the approximate nature of EMODPS, the diagnostic benchmarking uses SDP to evaluate the overall quality of the attained Pareto-approximate results. Although the Hoa Binh test case's relative simplicity should maximize the potential value of SDP, the results demonstrate that EMODPS successfully dominates the solutions derived via SDP. (C) 2015 America
暂无评论