Jakub Janostik

PSO as Complex Network—Capturing the Inner Dynamics—Initial Study

This paper presents an initial proposal of methodology for converting the inner dynamics of PSO algorithm into complex network. The motivation is in the recent trend of adaptive methods for improving the performance of evolutionary computational techniques. It seems very likely that the complex network and its statistical characteristics can be used within those adaptive approaches. The methodology described in this paper manages to put significant amount of information about the inner dynamics of PSO algorithm into a complex network.

Particle Swarm Optimizer with Diversity Measure Based on Swarm Representation in Complex Network

In this paper a alternative approach to the diversity guided particle swarm optimization (PSO) is investigated. The PSO shows acceptable performance on well-known test problems, however tends to suffer from premature convergence on multi-modal test problems. This premature convergence can be avoided by increasing diversity in search space. In this paper we introduce diversity measure based on graph representation of swam evolution and we discuss possibilities of graph representation of swarm population in adaptive control of PSO algorithm. Based on our findings we concluded, that network representation of evolution population and its subsequent analysis can be used in adaptive control, in various degrees of success.

Capturing Inner Dynamics of Firefly Algorithm in Complex Network—Initial Study

In this paper the idea of capturing inner dynamics of evolutionary process in complex network is discussed. Complex network, in itself, can contain large number of information about running evolutionary process. Consequently this information can be used in subsequent adaptive control of algorithm. We present initial study of methodology for capturing the information about algorithm and present possible uses for created networks. For this study the Firefly algorithm was selected.

Study on the Time Development of Complex Network for Metaheuristic

This work deals with the hybridization of the complex networks framework and evolutionary algorithms. The population is visualized as an evolving complex network, which exhibits non-trivial features. This paper investigates briefly the time development of complex network within the run of selected metaheuristic algorithm, which is Differential Evolution (DE). This paper also briefly discuss possible utilization of the complex network attributes such as adjacency graph, centralities, clustering coefficient and others. Experiments were performed for one selected DE strategy and one simple test function.

Preliminary Study on the Randomization and Sequencing for the Chaos Embedded Heuristic

This research deals with the hybridization of the two softcomputing fields, which are chaos theory and evolutionary computation. This paper investigates the utilization of the time-continuous chaotic system, which is UEDA oscillator, as the chaotic pseudo random number generator. (CPRNG). Repeated simulations were performed investigating the influence of the oscillator sampling time to the selected heuristic, which is differential evolution algorithm (DE). Through the utilization of time-continuous systems and with different sampling times from very small to bigger, it is possible to fully keep, suppress or remove the hidden complex chaotic dynamics from the generated data series. Experiments are focused on the preliminary investigation, whether the different randomization given by particular CPRNG or hidden complex chaotic dynamics providing the unique sequencing are beneficial to the heuristic performance. Initial experiments were performed on the selected test function in several dimension settings.

Study On Swarm Dynamics Converted Into Complex Network.

In this study it is presented a summarization of our research of possible ways of creating of complex networks from the inner dynamics of Swarm Intelligence based algorithms. The particle swarm optimization algorithm and the firefly algorithm are studied in this paper. Several methods of complex network creation are proposed and discussed alongside with possibilities for future research and application.

On the influence of different randomization and complex network analysis for differential evolution

This research deals with the hybridization of the chaos driven heuristics concept and complex networks framework for evolutionary algorithms. This paper aims on the experimental investigations on the influence of different randomization types for chaos-driven Differential Evolution (DE) through the analysis of complex network as a record of population dynamics. The population is visualized as an evolving complex network, which exhibits non-trivial features. Complex network attributes such as adjacency graph gives interconnectivity, centralities give the overview of convergence and stagnation, clustering coefficient gives diversity of population whereas other attributes like network density, average number of neighbors within the population shows efficiency of the network. Experiments were performed for two different DE strategies, four different randomization types and two simple test functions.

Converting PSO dynamics into complex network - Initial study

In this paper it is presented the initial study on the possibility of capturing the inner dynamic of Particle Swarm Optimization algorithm into a complex network structure. Inspired in previous works there are two different approaches for creating the complex network presented in this paper. Visualizations of the networks are presented and commented. The possibilities for future applications of the proposed design are given in detail.

Hybridization of adaptivity and chaotic dynamics for differential evolution

This research deals with the hybridization of the two modern approaches for evolutionary algorithms, which are the adaptivity and complex chaotic dynamics. This paper aims on the investigations on the chaos-driven adaptive Differential Evolution (DE) concept. This paper is aimed at the embedding of discrete dissipative chaotic systems in the form of chaotic pseudo random number generators for the state of the art adaptive representative jDE. Repeated simulations for two different driving chaotic systems were performed on the IEEE CEC 13 benchmark set. Finally, the obtained results are compared with the canonical not-chaotic jDE.

New Adaptive Approach for Multi-chaotic Differential Evolution Concept

This research deals with the hybridization of the two soft computing fields, which are the chaos theory and evolutionary computation. This paper aims on the investigations on the adaptive multi-chaos-driven evolutionary algorithm Differential Evolution (DE) concept. This paper is aimed at the embedding and adaptive alternating of set of two discrete dissipative chaotic systems in the form of chaotic pseudo random number generators for the DE. In this paper the novel adaptive concept of DE/rand/1/bin strategy driven alternately by two chaotic maps (systems) is introduced. From the previous research, it follows that very promising results were obtained through the utilization of different chaotic maps, which have unique properties with connection to DE. The idea is then to connect these two different influences to the performance of DE into the one adaptive multi-chaotic concept with automatic switching without prior knowledge of the optimization problem and without any manual setting of the “switching point”. Repeated simulations were performed on the IEEE CEC 13 benchmark set. Finally, the obtained results are compared with state of the art adaptive representative jDE.