Multi-Robot Task Allocation (MRTA) addresses the problems related to an efficient job assignment in a team of robots. This paper expresses MRTA as a generalization of the Multiple Traveling Salesman Problem (MTSP) and...
详细信息
ISBN:
(纸本)9789897582202
Multi-Robot Task Allocation (MRTA) addresses the problems related to an efficient job assignment in a team of robots. This paper expresses MRTA as a generalization of the Multiple Traveling Salesman Problem (MTSP) and utilizes evolutionary algorithms (EA) for optimal task assignment. The MTSP version of the problem is also solved using combinatorial optimization techniques and results are compared to demonstrate that EA can be effectively used for providing solutions to such problems.
Most of the existing community detection ( CD) methods are designed primarily for unsigned networks containing only positive links. Therefore, it is significant to explore and design effective CD methods for signed so...
详细信息
ISBN:
(纸本)9781538632215
Most of the existing community detection ( CD) methods are designed primarily for unsigned networks containing only positive links. Therefore, it is significant to explore and design effective CD methods for signed social networks (SNs) with both positive and negative links. In this paper, we first utilize decomposable characteristic of modularity Q to establish a bi-objective model for community detection from SNs. Afterwards, a conical area evolutionary algorithm based on the modularity Q (CAEAq-SN) is developed to solve this bi-objective model efficiently. Furthermore, a new tournament selection mechanism based on Q is applied to accelerate the convergence of Q. Experimental results on both benchmark networks and synthetic SNs indicate that CAEAq-SN achieves not only better community structures in term of both Q and NMI but also stronger robustness than the existing algorithm MEAs-SN.
This paper presents a verification of universal method for discretization of decision space in optimization algorithms. Real-world optimization tasks frequently use discontinuous decision variables and in order to eff...
详细信息
This paper presents a verification of universal method for discretization of decision space in optimization algorithms. Real-world optimization tasks frequently use discontinuous decision variables and in order to effectively optimize such tasks, it is necessary to exploit an optimization algorithm that meets such requirement. Unfortunately, very few evolutionary algorithms can naturally work with discontinuous decision space. The method that entitles all optimization algorithms to effectively solve problems with discrete variables is here described and experimentally verified.
Instance based classifiers, such as k-Nearest Neighbors, predict the class value of a new observation based on some distance or similarity measure between the new instance and the stored training data. However, due to...
详细信息
ISBN:
(纸本)9783319465623;9783319465616
Instance based classifiers, such as k-Nearest Neighbors, predict the class value of a new observation based on some distance or similarity measure between the new instance and the stored training data. However, due to the required distance calculations, classifying new instances becomes computationally expensive as the number of training observations increases. Therefore, instance selection techniques have been proposed to improve instance based classifiers by reducing the number of training instances that must be stored to achieve adequate classification rates. Although other methods exist, an evolutionary algorithm has been used for instance selection with some of the best results in regard to data reduction and preservation of classification accuracy. Unfortunately, the performance of the evolutionary algorithm for instance selection comes at the cost of longer computation times in comparison to classic instance selection techniques. In this work we introduce a new stopping criterion for the evolutionary algorithm which depends on the convergence of its fitness function. Experimentation shows that the new criterion results in less computation time while achieving comparable performance.
The identification of a nonlinear model often involves a significant amount of user interaction. The proposed SADE evolutionary algorithm-based identification approach for block-structured systems reduces this user in...
详细信息
ISBN:
(纸本)9783319549309;9783319549293
The identification of a nonlinear model often involves a significant amount of user interaction. The proposed SADE evolutionary algorithm-based identification approach for block-structured systems reduces this user interaction to a minimum. This is illustrated in this paper for the Wiener-Hammerstein class of systems. On top of this, most of the assumptions and limitations on the considered Wiener-Hammerstein system class can be omitted compared to the popular BLA and correlation based approaches. The developed identification algorithm is applied on the 2009 Wiener-Hammerstein benchmark to illustrate its good performance.
evolutionary algorithms have recently been used to create a wide range of artistic work. In this paper, we propose a new approach for the composition of new images from existing ones, that retain some salient features...
详细信息
ISBN:
(纸本)9781450349208
evolutionary algorithms have recently been used to create a wide range of artistic work. In this paper, we propose a new approach for the composition of new images from existing ones, that retain some salient features of the original images. We introduce evolutionary algorithms that create new images based on a fitness function that incorporates feature covariance matrices associated with different parts of the images. This approach is very flexible in that it can work with a wide range of features and enables targeting specific regions in the images. For the creation of the new images, we propose a population-based evolutionary algorithm with mutation and crossover operators based on random walks. Our experimental results reveal a spectrum of aesthetically pleasing images that can be obtained with the aid of our evolutionary process.
In the paper, we investigate the speeding up of the evolutionary induction of decision trees, which is an emerging alternative to greedy top-down solutions. In particular, we design and implement graphics processing u...
详细信息
ISBN:
(数字)9783319710693
ISBN:
(纸本)9783319710693;9783319710686
In the paper, we investigate the speeding up of the evolutionary induction of decision trees, which is an emerging alternative to greedy top-down solutions. In particular, we design and implement graphics processing units (GPU)-based parallelization to generate regression trees (decision trees employed to solve regression problems) on large-scale data. The most time consuming part of the algorithm, which is parallelized, is the evaluation of individuals in the population. Other parts of the algorithms (like selection, genetic operators) are performed sequentially on a CPU. A data-parallel approach is applied to split the dataset over the GPU cores. After each assigned chunk of data is processed, the results calculated on all GPU cores are merged and sent to the CPU. We use a Compute Unified Device Architecture (CUDA) programming model, which supports general purpose computation on a GPU (GPGPU). Experimental validation of the proposed approach is performed on artificial and real-life datasets. A computational performance comparison with the traditional CPU version shows that GPU-accelerated evolutionary induction of regression trees is significantly (even up to 1000 times) faster and allows for processing of much larger datasets.
The problem of community detection in complex networks is of high interest in many application domains including sociology, biology, mathematics and economy. Given a set of nodes and links between them, the aim of the...
详细信息
ISBN:
(纸本)9781538633687
The problem of community detection in complex networks is of high interest in many application domains including sociology, biology, mathematics and economy. Given a set of nodes and links between them, the aim of the problem is to find a grouping of nodes such that a strong community has dense intra-connections and sparse outside community links. In this paper, a coarse-grained evolutionary algorithm (EA) is developed to address this challenging problem. Several populations of potential solutions are evolved in parallel in an island model and periodically exchange certain individuals. Each population can be evolved by a different fitness function and several approaches to evaluate the community structure are considered in the current paper. Experiments are performed for real-world complex networks and results are analysed based on the normalized mutual information between the detected and the known community structure. Comparisons with the standard version of the EA based on different fitness functions are performed and the results confirm a good performance of the parallel EA in terms of solution quality and computational time.
Machine learning models exhibit vulnerability to adversarial examples i.e., artificially created inputs that become misinterpreted. The goal of this work is to explore black-box adversarial attacks on deep networks pe...
详细信息
Machine learning models exhibit vulnerability to adversarial examples i.e., artificially created inputs that become misinterpreted. The goal of this work is to explore black-box adversarial attacks on deep networks performing image classification. The role of surrogate machine learning models for adversarial attacks is studied, and a special version of a genetic algorithm for generating adversarial examples is proposed. The efficiency of attacks is validated by a multitude of experiments with the Fashion MNIST data set. The experimental results verify the usability of our approach with surprisingly good performance in several cases, such as non-targeted attack on residual networks.
Taking a multi-objective optimization approach to deformable image registration has recently gained attention, because such an approach removes the requirement of manually tuning the weights of all the involved object...
详细信息
ISBN:
(数字)9781510607125
ISBN:
(纸本)9781510607118;9781510607125
Taking a multi-objective optimization approach to deformable image registration has recently gained attention, because such an approach removes the requirement of manually tuning the weights of all the involved objectives. Especially for problems that require large complex deformations, this is a non-trivial task. From the resulting Pareto set of solutions one can then much more insightfully select a registration outcome that is most suitable for the problem at hand. To serve as an internal optimization engine, currently used multi-objective algorithms are competent, but rather inefficient. In this paper we largely improve upon this by introducing a multi-objective real-valued adaptation of the recently introduced Gene-pool Optimal Mixing evolutionary Algorithm (GOMEA) for discrete optimization. In this work, GOMEA is tailored specifically to the problem of deformable image registration to obtain substantially improved efficiency. This improvement is achieved by exploiting a key strength of GOMEA: iteratively improving small parts of solutions, allowing to faster exploit the impact of such updates on the objectives at hand through partial evaluations. We performed experiments on three registration problems. In particular, an artificial problem containing a disappearing structure, a pair of pre- and post-operative breast CT scans, and a pair of breast MRI scans acquired in prone and supine position were considered. Results show that compared to the previously used evolutionary algorithm, GOMEA obtains a speed-up of up to a factor of similar to 1600 on the tested registration problems while achieving registration outcomes of similar quality.
暂无评论