The range of applications of Neural Networks encompasses image classification. However, Neural Networks are vulnerable to attacks, and may misclassify adversarial images, leading to potentially disastrous consequences...
详细信息
The range of applications of Neural Networks encompasses image classification. However, Neural Networks are vulnerable to attacks, and may misclassify adversarial images, leading to potentially disastrous consequences. Pursuing some of our previous work, we provide an extended proof of concept of a black-box, targeted, non-parametric attack using evolutionary algorithms to fool both Neural Networks and humans at the task of image classification. Our feasibility study is performed on VGG-16 trained on CIFAR-10. For any category c(A) of CIFAR-10, one chooses an image A classified by VGG-16 as belonging to c(A) . From there, two scenarios are addressed. In the first scenario, a target category c(t) not equal c(A) is fixed a priori. We construct an evolutionary algorithm that evolves A to a modified image that VGG-16 classifies as belonging to c(t) . In the second scenario, we construct another evolutionary algorithm that evolves A to a modified image that VGG-16 is unable to classify. In both scenarios, the obtained adversarial images remain so close to the original one that a human would likely classify them as still belonging to A .
An introduction to the special issue of the journal is presented in which the editor discusses the trends of using evolutionary algorithms (EAs) in supply chain management (SCM), the role of Pareto analysis in EAs, an...
详细信息
An introduction to the special issue of the journal is presented in which the editor discusses the trends of using evolutionary algorithms (EAs) in supply chain management (SCM), the role of Pareto analysis in EAs, and multi-period stochastic modelling frameworks in fashion products.
Over the past 30 years, algorithms that model natural evolution have generated robust search methods. These so-called evolutionary algorithms have been successfully applied to a wide range of problems. This paper disc...
详细信息
Over the past 30 years, algorithms that model natural evolution have generated robust search methods. These so-called evolutionary algorithms have been successfully applied to a wide range of problems. This paper discusses two types of evolutionary algorithms and their application to a problem in shape representation. Genetic algorithms and evolutionary programming, although both based on evolutionary principles, each place different emphasis on what drives the evolutionary process. While genetic algorithms rely on mimicking specific genotypic transformations, evolutionary programming emphasizes phenotypic adaptation. Results presented show the success of evolutionary programming in solving an example of a fractal inverse problem, but indicate that a genetic algorithm is not as successful. Reasons for this disparity are discussed.
In this article, we discuss the progress achieved with the use of evolutionary algorithms for the analysis of H-1 Nuclear Magnetic Resonance spectra of solutes in orientationally ordered liquids. With these tools the ...
详细信息
In this article, we discuss the progress achieved with the use of evolutionary algorithms for the analysis of H-1 Nuclear Magnetic Resonance spectra of solutes in orientationally ordered liquids. With these tools the analysis of extremely complex spectra that were hitherto impossible to solve has now become eminently feasible. We discuss applications to 2 molecules of special interest: (a) hexamethylbenzene, which is a text book example of steric hindrance between adjacent rotating methyl groups;and (b) cyclohexane which is the standard example of interconversion between various molecular conformations. New interesting physics is obtained in both cases.
This paper wishes to describe evolutionary algorithms as an effective means for the solution of the Aerofoil Design Optimisation in Aerodynamics. Firstly the basic ideas underlying evolutionary algorithms are outlined...
详细信息
This paper wishes to describe evolutionary algorithms as an effective means for the solution of the Aerofoil Design Optimisation in Aerodynamics. Firstly the basic ideas underlying evolutionary algorithms are outlined. Several versions of evolutionary algorithms are briefly described, focussing on their similarities and on their differences as well. Then their application to both Direct and Inverse Aerofoil Design Problem is described, and results are given. Finally, several possible parallel models for evolutionary algorithms are discussed, and the results of the application of one of them to the above problem are presented.
One of the goals of computational chemistry is the automated de novo design of bioactive molecules. Despite significant progress in computational approaches to ligand design and efficient evaluation of binding energy,...
详细信息
One of the goals of computational chemistry is the automated de novo design of bioactive molecules. Despite significant progress in computational approaches to ligand design and efficient evaluation of binding energy, novel procedures for ligand design are required. evolutionary computation provides a new approach to this design issue. This paper presents an automated methodology for computer-aided peptide design based on evolutionary algorithms. It provides an automatic tool for peptide de novo design, based on protein surface patches defined by user. Regarding the restrictive constrains of this problem a special emphasis has been made on the design of the evolutionary algorithms implemented.
In the last years, several real-world problems that require to optimise an increasing number of variables have appeared. This type of optimisation, called large-scale global optimisation, is hard due to the huge incre...
详细信息
In the last years, several real-world problems that require to optimise an increasing number of variables have appeared. This type of optimisation, called large-scale global optimisation, is hard due to the huge increase of the domain search due to the dimensionality. Large-scale global optimisation is a research area getting more attention in the last years, thus many algorithms, mainly evolutionary algorithms, have been specially designed to tackle it. In this paper, we give a brief introduction of several of them and their techniques, remarking techniques based on grouping of variables and memetic algorithms, because they are two promising approaches. Also, we have reviewed the winners of the different competitions in the area, to give a snapshot of the algorithms that have obtained the best results in this area. Finally, several interesting trends in the research area have been pointed out, and some future trends and challenges have been suggested.
We analyze the performance of evolutionary algorithms on various matroid optimization problems that encompass a vast number of efficiently solvable as well as NP-hard combinatorial optimization problems (including man...
详细信息
We analyze the performance of evolutionary algorithms on various matroid optimization problems that encompass a vast number of efficiently solvable as well as NP-hard combinatorial optimization problems (including many well-known examples such as minimum spanning tree and maximum bipartite matching). We obtain very promising bounds on the expected running time and quality of the computed solution. Our results establish a better theoretical understanding of why randomized search heuristics yield empirically good results for many real-world optimization problems.
This paper considers the problem of scheduling part families and jobs within each part family in a flowshop manufacturing cell with sequence dependent family setups times where it is desired to minimize the makespan w...
详细信息
This paper considers the problem of scheduling part families and jobs within each part family in a flowshop manufacturing cell with sequence dependent family setups times where it is desired to minimize the makespan while processing parts (jobs) in each family together. Two evolutionary algorithms-a Genetic Algorithm and a Memetic Algorithm with local search-are proposed and empirically evaluated as to their effectiveness in finding optimal permutation schedules. The proposed algorithms use a compact representation for the solution and a hierarchically structured population where the number of possible neighborhoods is limited by dividing the population into clusters. In comparison to a Multi-Start procedure, solutions obtained by the proposed evolutionary algorithms were very close to the lower bounds for all problem instances. Moreover, the comparison against the previous best algorithm, a heuristic named CMD, indicated a considerable performance improvement. (c) 2005 Elsevier Ltd. All rights reserved.
In this paper the method of selecting a representative subset of Pareto optimal solutions is used to make the search of Pareto frontier more effective. Firstly, the evolutionary algorithm method for generating a set o...
详细信息
ISBN:
(纸本)3540417451
In this paper the method of selecting a representative subset of Pareto optimal solutions is used to make the search of Pareto frontier more effective. Firstly, the evolutionary algorithm method for generating a set of Pareto optimal solutions is described. Then, indiscernibility interval method is applied to select representative subset of Pareto optimal solutions. The main idea of this method consists in removing from the set of Pareto optimal solutions these solutions, which are close to each other in the space of objectives, i.e., those solutions for which the values of the objective functions differ less than an indiscernibility interval. The set of Pareto optimal solutions is reduced using indiscernibility interval method after running a certain number of generations. This process can be called the filtration process in which less important Pareto optimal solutions are removed from the existing set. Finally, two design optimization problems are solved using the proposed method. From these examples it is clear that the computation time can be reduced significantly and still the real Pareto frontier obtained.
暂无评论