Simultaneous Optimization of Weights and Structure of an RBF Neural Network

Lefort, Virginie; Knibbe, Carole; Beslon, Guillaume; Favrel, Joël

doi:10.1007/11740698_5

Cited by 4 publications

(3 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…model complexity, representation ability, and model smoothness, and RBF Network ensemble has been constructed of this Pareto set. A new evolutionary algorithm, the RBF-Gene algorithm, has been applied to optimize RBF Networks [8]. Unlike other works, their algorithm can evolve both from the structure and the numerical parameters of the network: it is able to evolve the number of neurons and their weights.…”

Section: Related Work Of Evolutionary Multi-objective Rbf Networkmentioning

confidence: 99%

Improving generalization of radial basis function network with adaptive multi-objective particle swarm optimization

Qasem

Shamsuddin

2009

2009 IEEE International Conference on Systems, Man and Cybernetics

View full text Add to dashboard Cite

In this paper, an adaptive evolutionary multiobjective selection method of RBF Networks structure is discussed. The candidates of RBF Network structures are encoded into particles in Particle Swarm Optimization (PSO). These particles evolve toward Pareto-optimal front defined by several objective functions with model accuracy and complexity. The problem of unsupervised and supervised learning is discussed with Adaptive Multi-Objective PSO (AMOPSO). This study suggests an approach of RBF Network training through simultaneous optimization of architectures and weights withAdaptive PSO-based multi-objective algorithm. Our goal is to determine whether Adaptive Multi-objective PSO can train RBF Networks, and the performance is validated on accuracy and complexity. The experiments are conducted on two benchmark datasets obtained from the machine learning repository. The results show that our proposed method provides an effective means for training RBF Networks that is competitive with PSObased multi-objective algorithm.

show abstract

Section: Related Work Of Evolutionary Multi-objective Rbf Networkmentioning

confidence: 99%

Improving generalization of radial basis function network with adaptive multi-objective particle swarm optimization

Qasem

Shamsuddin

2009

2009 IEEE International Conference on Systems, Man and Cybernetics

View full text Add to dashboard Cite

show abstract

“…This trade-off is a well-known problem as the Multiobjective Optimization Problem (MOP) [49,50,51,52]. This paper applies NSGA II (Non-dominated Sorting Genetic Algorithm) proposed by Deb et al (2002) to solve this problem, as it has recently been frequently applied to various scenarios [53,54,55,56]. On the other hand, for (near) optimal estimation and adjustment of two others RBF parameters (units' widths and output weights), we implement Particle Swarm Optimization (PSO) that favors global and local search of its interacting particles which has proved to be effective in finding the optimum in a search space [57,58,59].…”

Section: Introductionmentioning

confidence: 99%

A hybrid multiobjective RBF-PSO method for mitigating DoS attacks in Named Data Networking

Karami

Zapata

2015

Neurocomputing

View full text Add to dashboard Cite

Named Data Networking (NDN) is a promising network architecture being considered as a possible replacement for the current IP-based (host-centric) Internet infrastructure. NDN can overcome the fundamental limitations of the current Internet, in particular, Denial-of-Service (DoS) attacks. However, NDN can be subject to new type of DoS attacks namely Interest flooding attacks and content poisoning. These types of attacks exploit key architectural features of NDN. This paper presents a new intelligent hybrid algorithm for proactive detection of DoS attacks and adaptive mitigation reaction in NDN. In the detection phase, a combination of multiobjective evolutionary optimization algorithm with PSO in the context of the RBF neural network has been applied in order to improve the accuracy of DoS attack prediction. Performance of the proposed hybrid approach is also evaluated successfully by some benchmark problems. In the adaptive reaction phase, we introduced a framework for mitigating DoS attacks based on the misbehaving type of network nodes. The evaluation through simulations shows that the proposed intelligent hybrid algorithm (proactive detection and adaptive reaction) can quickly and effectively respond and mitigate DoS attacks in adverse conditions in terms of the applied performance criteria.

show abstract

“…This trade-off is a well-known problem as the Multiobjective Optimization Problem (MOP) [150,169,170,171]. This dissertation applies NSGA II (Non-dominated Sorting Genetic Algorithm) proposed by Deb et al (2002) to solve this problem, as it has recently been frequently applied to various scenarios [172,173,174,175]. On the other hand, for (near) optimal estimation and adjustment of two others RBF parameters (units' widths and output weights), we implement Particle Swarm Optimization (PSO) that favors global and local search of its interacting particles which has proved to be effective in finding the optimum in a search space [117,118,176].…”

Section: Methods For Mitigating Dos Attacksmentioning

confidence: 99%

The use of computational intelligence for security in named data networking

Karami

View full text Add to dashboard Cite

Information-Centric Networking (ICN) has recently been considered as a promising paradigm for the next-generation Internet, shifting from the sender-driven end-to-end communication paradigma to a receiver-driven content retrieval paradigm. In ICN, content -rather than hosts, like in IP-based design- plays the central role in the communications. This change from host-centric to content-centric has several significant advantages such as network load reduction, low dissemination latency, scalability, etc. One of the main design requirements for the ICN architectures -since the beginning of their design- has been strong security. Named Data Networking (NDN) (also referred to as Content-Centric Networking (CCN) or Data-Centric Networking (DCN)) is one of these architectures that are the focus of an ongoing research effort that aims to become the way Internet will operate in the future. Existing research into security of NDN is at an early stage and many designs are still incomplete. To make NDN a fully working system at Internet scale, there are still many missing pieces to be filled in. In this dissertation, we study the four most important security issues in NDN in order to defense against new forms of -potentially unknown- attacks, ensure privacy, achieve high availability, and block malicious network traffics belonging to attackers or at least limit their effectiveness, i.e., anomaly detection, DoS/DDoS attacks, congestion control, and cache pollution attacks. In order to protect NDN infrastructure, we need flexible, adaptable and robust defense systems which can make intelligent -and real-time- decisions to enable network entities to behave in an adaptive and intelligent manner. In this context, the characteristics of Computational Intelligence (CI) methods such as adaption, fault tolerance, high computational speed and error resilient against noisy information, make them suitable to be applied to the problem of NDN security, which can highlight promising new research directions. Hence, we suggest new hybrid CI-based methods to make NDN a more reliable and viable architecture for the future Internet. Information-Centric Networking (ICN) ha sido recientemente considerado como un paradigma prometedor parala nueva generación de Internet, pasando del paradigma de la comunicación de extremo a extremo impulsada por el emisora un paradigma de obtención de contenidos impulsada por el receptor. En ICN, el contenido (más que los nodos, como sucede en redes IPactuales) juega el papel central en las comunicaciones. Este cambio de "host-centric" a "content-centric" tiene varias ventajas importantes como la reducción de la carga de red, la baja latencia, escalabilidad, etc. Uno de los principales requisitos de diseño para las arquitecturas ICN (ya desde el principiode su diseño) ha sido una fuerte seguridad. Named Data Networking (NDN) (también conocida como Content-Centric Networking (CCN) o Data-Centric Networking (DCN)) es una de estas arquitecturas que son objetode investigación y que tiene como objetivo convertirse en la forma en que Internet funcionará en el futuro. Laseguridad de NDN está aún en una etapa inicial. Para hacer NDN un sistema totalmente funcional a escala de Internet, todavía hay muchas piezas que faltan por diseñar. Enesta tesis, estudiamos los cuatro problemas de seguridad más importantes de NDN, para defendersecontra nuevas formas de ataques (incluyendo los potencialmente desconocidos), asegurar la privacidad, lograr una alta disponibilidad, y bloquear los tráficos de red maliciosos o al menos limitar su eficacia. Estos cuatro problemas son: detección de anomalías, ataques DoS / DDoS, control de congestión y ataques de contaminación caché. Para solventar tales problemas necesitamos sistemas de defensa flexibles, adaptables y robustos que puedantomar decisiones inteligentes en tiempo real para permitir a las entidades de red que se comporten de manera rápida e inteligente. Es por ello que utilizamos Inteligencia Computacional (IC), ya que sus características (la adaptación, la tolerancia a fallos, alta velocidad de cálculo y funcionamiento adecuado con información con altos niveles de ruido), la hace adecuada para ser aplicada al problema de la seguridad NDN

show abstract

Simultaneous Optimization of Weights and Structure of an RBF Neural Network

Cited by 4 publications

References 9 publications

Improving generalization of radial basis function network with adaptive multi-objective particle swarm optimization

Improving generalization of radial basis function network with adaptive multi-objective particle swarm optimization

A hybrid multiobjective RBF-PSO method for mitigating DoS attacks in Named Data Networking

The use of computational intelligence for security in named data networking

Contact Info

Product

Resources

About