Search

Your search keyword '"Evolution Strategie"' showing total 6 results
Start Over Descriptor "Evolution Strategie"
6 results on '"Evolution Strategie"'

1. Evolutionary Approaches for Adversarial Attacks on Neural Source Code Classifiers

Abstract

As the prevalence and sophistication of cyber threats continue to increase, the development of robust vulnerability detection techniques becomes paramount in ensuring the security of computer systems. Neural models have demonstrated significant potential in identifying vulnerabilities; however, they are not immune to adversarial attacks. This paper presents a set of evolutionary techniques for generating adversarial instances to enhance the resilience of neural models used for vulnerability detection. The proposed approaches leverage an evolution strategy (ES) algorithm that utilizes as the fitness function the output of the neural network to deceive. By starting from existing instances, the algorithm evolves individuals, represented by source code snippets, by applying semantic-preserving transformations, while utilizing the fitness to invert their original classification. This iterative process facilitates the generation of adversarial instances that can mislead the vulnerability detection models while maintaining the original behavior of the source code. The significance of this research lies in its contribution to the field of cybersecurity by addressing the need for enhanced resilience against adversarial attacks in vulnerability detection models. The evolutionary approach provides a systematic framework for generating adversarial instances, allowing for the identification and mitigation of weaknesses in AI classifiers.

Published
2023

2. Hyper-bent Boolean Functions and Evolutionary Algorithms

Abstract

Bent Boolean functions play an important role in the design of secure symmetric ciphers, since they achieve the maximum distance from affine functions allowed by Parseval’s relation. Hyper-bent functions, in turn, are those bent functions which additionally reach maximum distance from all bijective monomial functions, and provide further security towards approximation attacks. Being characterized by a stricter definition, hyper-bent functions are rarer than bent functions, and much more difficult to construct. In this paper, we employ several evolutionary algorithms in order to evolve hyper-bent Boolean functions of various sizes. Our results show that hyper-bent functions are extremely difficult to evolve, since we manage to find such functions only for the smallest investigated size. Interestingly, we are able to identify this difficulty as not lying in the evolution of hyper-bent functions itself, but rather in evolving some of their components, i.e. bent functions. Finally, we present an additional parameter to evaluate the performance of evolutionary algorithms when evolving Boolean functions: the diversity of the obtained solutions.

Published
2019

3. Hyper-bent Boolean Functions and Evolutionary Algorithms

Author

Domagoj Jakobovic, Luca Mariot, Alberto Leporati, Stjepan Picek, Mariot, L, Jakobovic, D, Leporati, A, and Picek, S

Subjects
Discrete mathematics, 0209 industrial biotechnology, Monomial, Bent function, Computer science, Bent molecular geometry, Evolutionary algorithm, INF/01 - INFORMATICA, Genetic programming, 02 engineering and technology, Evolution strategie, Bent functions, Hyper-bent functions, Genetic algorithms, Evolution strategies, GeneralLiterature_MISCELLANEOUS, Parseval's theorem, 020901 industrial engineering & automation, Genetic algorithm, 0202 electrical engineering, electronic engineering, information engineering, Bijection, Physics::Accelerator Physics, 020201 artificial intelligence & image processing, Affine transformation, Boolean function
Abstract

Bent Boolean functions play an important role in the design of secure symmetric ciphers, since they achieve the maximum distance from affine functions allowed by Parseval’s relation. Hyper-bent functions, in turn, are those bent functions which additionally reach maximum distance from all bijective monomial functions, and provide further security towards approximation attacks. Being characterized by a stricter definition, hyper-bent functions are rarer than bent functions, and much more difficult to construct. In this paper, we employ several evolutionary algorithms in order to evolve hyper-bent Boolean functions of various sizes. Our results show that hyper-bent functions are extremely difficult to evolve, since we manage to find such functions only for the smallest investigated size. Interestingly, we are able to identify this difficulty as not lying in the evolution of hyper-bent functions itself, but rather in evolving some of their components, i.e. bent functions. Finally, we present an additional parameter to evaluate the performance of evolutionary algorithms when evolving Boolean functions: the diversity of the obtained solutions.

Published
2019

4. Dynamical modeling and parameter identification of seismic isolation systems by evolution strategies

Author

De Felice, M. and De Felice, M.

Subjects
Real-world applications, Evolution Strategies, Structural system identification, Real-world application, Earthquake engineering, Evolution Strategie, CMA-ES
Abstract

An application of Evolution Strategies (ESs) to the dynamic identification of hybrid seismic isolation systems is presented. It is shown how ESs are highly effective for the optimisation of the test problem defined in previous work for methodology validation. The acceleration records of a number of dynamic tests performed on a seismically isolated building are used as reference data for the parameter identification. The application of CMA-ES to a previously existing model considerably improves previous results but at the same time reveals limitations of the model. To investigate the problem three new mechanical models with higher number of parameters are developed. The application of CMA-ES to the best designed model allows improvements of up to 79% compared to the solutions previously available in literature. © Springer-Verlag Berlin Heidelberg 2013.

Published
2013

5. Gestion de l'évolution dans les bases de connaissances : une approche par les règles

Author

Bounaas, Fethi, Laboratoire de Génie Informatique (LGI - IMAG), Université Joseph Fourier - Grenoble 1 (UJF)-IMAG, Institut National Polytechnique de Grenoble - INPG, and Nguyen Gia Toan

Subjects
représentation de connaissances, active rules, évolution de schémas, knowledge representation, schema evolution, instance classification, règles actives, stratégie d'évolution, [INFO.INFO-HC]Computer Science [cs]/Human-Computer Interaction [cs.HC], Dynamique des objets, classification d'instances, Dynamics of objects, evolution strategie
Abstract

The work presented in this thesis concerns schema and object evolution in knowledge based systems. It is realized in the framework of the knowledge representation system Shood. We try to solve extensibility and reuse problems for evolution mechanisms. We propose an evolution system which allows the definition and the management of the dynamics. This management is realized by a set of mechanisms, such as instance classification, active rules, and a set of modification operations called the evolution support. We propose also a mechanism based on evolution rules and strategies to allow the declarative expression of constraints for structures and data evolution. The designer can define strategies to organize these constraints. According to his needs, he can define multiple strategies, but only one will be active at a given moment. The mechanism implementing evolution rules and strategies is developped mainly by mapping between an evolution rule and several active rules, and between a strategy and a rule base.; Le travail présenté dans cette thèse aborde les problèmes de gestion de l'évolution des schémas et des objets dans les systèmes à base de connaissances. Ce travail a été réalisé dans le cadre du système de représentation de connaissances objets SHOOD. Cette étude tente d'apporter des solutions aux problèmes d'extensibilité et de réutilisation des mécanismes d'évolution. Nous proposons un système d'évolution permettant la définition et la mise en ouvre de la dynamique. Cette mise en ouvre est réalisée par un ensemble de mécanismes tels que la classification d'instances et les règles actives ou règles ECA (Evenement,Condition, Action), ainsi qu'un ensemble d'opérations de manipulation : le support d'évolution. Nous proposons de même un mécanisme de règles et de stratégies d'évolution pour permettre une expression déclarative des contraintes d'évolution de structures ou de données. Le concepteur peut définir des stratégies pour regrouper ces contraintes. Suivant ses besoins, il peut faire cohabiter plusieurs stratégies, mais une seule sera active à un moment donné. Ce mécanisme de règles et de stratégies d'évolution est principalement développé en établissant une correspondance entre une règle d'évolution et plusieurs règles ECA, ainsi qu'une correspondance entre une stratégie et une base de règles ECA.

Published
1995

6. Evolution management in knowledge bases : an approach using rules

Author

Bounaas, Fethi, Laboratoire de Génie Informatique (LGI - IMAG), Université Joseph Fourier - Grenoble 1 (UJF)-IMAG, Institut National Polytechnique de Grenoble - INPG, and Nguyen Gia Toan

Subjects
représentation de connaissances, active rules, évolution de schémas, knowledge representation, schema evolution, instance classification, règles actives, stratégie d'évolution, [INFO.INFO-HC]Computer Science [cs]/Human-Computer Interaction [cs.HC], Dynamique des objets, classification d'instances, Dynamics of objects, evolution strategie
Abstract

The work presented in this thesis concerns schema and object evolution in knowledge based systems. It is realized in the framework of the knowledge representation system Shood. We try to solve extensibility and reuse problems for evolution mechanisms. We propose an evolution system which allows the definition and the management of the dynamics. This management is realized by a set of mechanisms, such as instance classification, active rules, and a set of modification operations called the evolution support. We propose also a mechanism based on evolution rules and strategies to allow the declarative expression of constraints for structures and data evolution. The designer can define strategies to organize these constraints. According to his needs, he can define multiple strategies, but only one will be active at a given moment. The mechanism implementing evolution rules and strategies is developped mainly by mapping between an evolution rule and several active rules, and between a strategy and a rule base.; Le travail présenté dans cette thèse aborde les problèmes de gestion de l'évolution des schémas et des objets dans les systèmes à base de connaissances. Ce travail a été réalisé dans le cadre du système de représentation de connaissances objets SHOOD. Cette étude tente d'apporter des solutions aux problèmes d'extensibilité et de réutilisation des mécanismes d'évolution. Nous proposons un système d'évolution permettant la définition et la mise en ouvre de la dynamique. Cette mise en ouvre est réalisée par un ensemble de mécanismes tels que la classification d'instances et les règles actives ou règles ECA (Evenement,Condition, Action), ainsi qu'un ensemble d'opérations de manipulation : le support d'évolution. Nous proposons de même un mécanisme de règles et de stratégies d'évolution pour permettre une expression déclarative des contraintes d'évolution de structures ou de données. Le concepteur peut définir des stratégies pour regrouper ces contraintes. Suivant ses besoins, il peut faire cohabiter plusieurs stratégies, mais une seule sera active à un moment donné. Ce mécanisme de règles et de stratégies d'évolution est principalement développé en établissant une correspondance entre une règle d'évolution et plusieurs règles ECA, ainsi qu'une correspondance entre une stratégie et une base de règles ECA.

Published
1995

Catalog

Books, media, physical & digital resources
See catalog results