Category: Evolutionary Intelligence

Collective neuro-evolution for evolving specialized sensor resolutions in a multi-rover task

Abstract  This article presents results from an evaluation of the collective neuro-evolution (CONE) controller design method. CONE solves
collective behavior tasks, and increases task performance via facilitating emergent behavioral speciali…

Abstract  

This article presents results from an evaluation of the collective neuro-evolution (CONE) controller design method. CONE solves
collective behavior tasks, and increases task performance via facilitating emergent behavioral specialization. Emergent specialization
is guided by genotype and behavioral specialization difference metrics that regulate genotype recombination. CONE is comparatively
tested and evaluated with similar neuro-evolution methods in an extension of the multi-rover task, where behavioral specialization
is known to benefit task performance. The task is for multiple simulated autonomous vehicles (rovers) to maximize the detection
of points of interest (red rocks) in a virtual environment. Results indicate that CONE is appropriate for deriving sets of
specialized rover behaviors that complement each other such that a higher task performance, comparative to related controller
design methods, is attained in the multi-rover task.

  • Content Type Journal Article
  • DOI 10.1007/s12065-009-0034-z
  • Authors
    • G. S. Nitschke, Computational Intelligence Research Group, Department of Computer Science, University of Pretoria, Pretoria, 0002 South Africa
    • M. C. Schut, Computational Intelligence Group, Vrije Universiteit Amsterdam, De Boelelaan 1081a, 1081 HV Amsterdam, The Netherlands
    • A. E. Eiben, Computational Intelligence Group, Vrije Universiteit Amsterdam, De Boelelaan 1081a, 1081 HV Amsterdam, The Netherlands

Adaptive ε-Ranking on many-objective problems

Abstract  This work proposes Adaptive ε-Ranking to enhance Pareto based selection, aiming to develop effective many-objective evolutionary optimization algorithms. ε-Ranking fine grains ranking of solutions after they have been ranked by
P…

Abstract  This work proposes Adaptive ε-Ranking to enhance Pareto based selection, aiming to develop effective many-objective evolutionary optimization algorithms. ε-Ranking fine grains ranking of solutions after they have been ranked by
Pareto dominance, using a randomized sampling procedure combined with ε-dominance to favor a good distribution of the samples.
In the proposed method, sampled solutions keep their initial rank and solutions located within the virtually expanded ε-dominance
regions of the sampled solutions are demoted to an inferior rank. The parameter ε that determines the expanded regions of
dominance of the sampled solutions is adapted at each generation so that the number of best-ranked solutions is kept close
to a desired number that is expressed as a fraction of the population size. We enhance NSGA-II with the proposed method and
analyze its performance on MNK-Landscapes, showing that the adaptive method works effectively and that compared to NSGA-II
convergence and diversity of solutions can be improved remarkably on MNK-Landscapes with 3 ≤ M ≤ 10 objectives. Also, we compare the performance of Adaptive ε-Ranking with two representative many-objective evolutionary
algorithms on DTLZ continuous functions. Results on DTLZ functions with 3 ≤ M ≤ 10 objectives suggest that the three many-objective approaches emphasize different areas of objective space and could be
used as complementary strategies to produce a better approximation of the Pareto front.

  • Content Type Journal Article
  • DOI 10.1007/s12065-009-0031-2
  • Authors
    • Hernán Aguirre, Shinshu University International Young Researcher Empowerment Center, Faculty of Engineering 4-17-1 Wakasato Nagano 380-8553 Japan
    • Kiyoshi Tanaka, Shinshu University Faculty of Engineering 4-17-1 Wakasato Nagano 380-8553 Japan

Special issue on simulated evolution and learning

Special issue on simulated evolution and learning
Content Type Journal ArticleDOI 10.1007/s12065-009-0033-0Authors
Michael Kirley, The University of Melbourne Department of Computer Science and Software Engineering Melbourne AustraliaMengjie Zhang, …

Special issue on simulated evolution and learning

  • Content Type Journal Article
  • DOI 10.1007/s12065-009-0033-0
  • Authors
    • Michael Kirley, The University of Melbourne Department of Computer Science and Software Engineering Melbourne Australia
    • Mengjie Zhang, Victoria University School of Engineering and Computer Science Wellington New Zealand
    • Xiaodong Li, RMIT University School of Computer Science and Information Technology Melbourne Australia

The effects of time-varying rewards on the evolution of cooperation

Abstract  Understanding how cooperative behavior emerges within a population of autonomous individuals has been the focus of a great
deal of research in biology, economics and more recently in the multi-agent systems domain. However, there a…

Abstract  Understanding how cooperative behavior emerges within a population of autonomous individuals has been the focus of a great
deal of research in biology, economics and more recently in the multi-agent systems domain. However, there are still many
open questions. In this paper, we address some of these questions by investigating the effects of time-varying, non-symmetric
rewards on the evolution of cooperation in the spatial Prisoner’s dilemma game. The rationale behind this approach is based
on the notion that the associated payoffs from pursuing certain strategies do vary among members of real-world populations.
In our model, agents with limited cognitive capacity play the game with their local neighbours. In addition to its game playing
strategy, each agent has additional attributes that can be used to control the number of rounds of the game the agent actually
participates in, as well as the magnitude of any rewards that it receives. Numerical simulations show that dynamic updates
to payoff values induce a change in equilibrium cooperation levels. This suggests that heterogeneous payoff values and social
diversity within a cost-benefit context are important factors in the promotion of cooperation in the spatial Prisoner’s dilemma
game.

  • Content Type Journal Article
  • DOI 10.1007/s12065-009-0032-1
  • Authors
    • Golriz Rezaei, The University of Melbourne Department of Computer Science and Software Engineering Melbourne VIC Australia
    • Michael Kirley, The University of Melbourne Department of Computer Science and Software Engineering Melbourne VIC Australia

Numerical simplification for bloat control and analysis of building blocks in genetic programming

Abstract  In tree-based genetic programming, there is a tendency for the size of the programs to increase from generation to generation,
a phenomenon known as bloat. It is standard practice to place some form of control on program size eithe…

Abstract  In tree-based genetic programming, there is a tendency for the size of the programs to increase from generation to generation,
a phenomenon known as bloat. It is standard practice to place some form of control on program size either by limiting the
number of nodes or the depth of the program trees, or by adding a component to the fitness function that rewards smaller programs
(parsimony pressure). Others have proposed directly simplifying individual programs using algebraic methods. In this paper,
we add node-based numerical simplification as a tree pruning criterion to control program size. We investigate the effect
of on-line program simplification, both algebraic and numerical, on program size and resource usage. We also investigate the
distribution of building blocks within a genetic programming population and how this is changed by using simplification. We
show that simplification results in reductions in expected program size, memory use and computation time. We also show that
numerical simplification performs at least as well as algebraic simplification, and in some cases will outperform algebraic
simplification. We further show that although the two on-line simplification methods destroy some existing building blocks,
they effectively generate new more diverse building blocks during evolution, which compensates for the negative effect of
disruption of building blocks.

  • Content Type Journal Article
  • DOI 10.1007/s12065-009-0029-9
  • Authors
    • David Kinzett, Victoria University of Wellington School of Engineering and Computer Science PO Box 600 Wellington New Zealand
    • Mark Johnston, Victoria University of Wellington School of Mathematics, Statistics and Operations Research PO Box 600 Wellington New Zealand
    • Mengjie Zhang, Victoria University of Wellington School of Engineering and Computer Science PO Box 600 Wellington New Zealand

Improving the performance of evolutionary algorithms in grid-based puzzles resolution

Abstract  This paper proposes several modifications to existing hybrid evolutionary algorithms in grid-based puzzles, using a-priori
probabilities of 0/1 occurrence in binary encodings. This calculation of a-priori probabilities of bits is p…

Abstract  This paper proposes several modifications to existing hybrid evolutionary algorithms in grid-based puzzles, using a-priori
probabilities of 0/1 occurrence in binary encodings. This calculation of a-priori probabilities of bits is possible in grid-based
problems (puzzles in this case) due to their special structure, with the solution confined into a grid. The work is focused
in two different grid-based puzzles, the Japanese puzzles and the Light-up puzzle, each one having special characteristics
in terms of constraints, which must be taken into account for the probabilities of bit calculation. For these puzzles, we
show the process of a-priori probabilities calculation, and we modify the initialization of the EAs to improve their performance.
We also include novel mutation operators based on a-priori probabilities, which makes more effective the evolutionary search
of the algorithms in the tackled puzzles. The performance of the algorithms with these new initialization and novel mutation
operators is compared with the performance without them. We show that the new initialization and operators based on a-priori
probabilities of bits make the evolutionary search more effective and also improve the scalability of the algorithms.

  • Content Type Journal Article
  • DOI 10.1007/s12065-009-0030-3
  • Authors
    • E. G. Ortiz-García, Universidad de Alcalá, Escuela Politécnica Superior Department of Signal Theory and Communications Alcalá de Henares 28871 Madrid Spain
    • S. Salcedo-Sanz, Universidad de Alcalá, Escuela Politécnica Superior Department of Signal Theory and Communications Alcalá de Henares 28871 Madrid Spain
    • Á. M. Pérez-Bellido, Universidad de Alcalá, Escuela Politécnica Superior Department of Signal Theory and Communications Alcalá de Henares 28871 Madrid Spain
    • L. Carro-Calvo, Universidad de Alcalá, Escuela Politécnica Superior Department of Signal Theory and Communications Alcalá de Henares 28871 Madrid Spain
    • A. Portilla-Figueras, Universidad de Alcalá, Escuela Politécnica Superior Department of Signal Theory and Communications Alcalá de Henares 28871 Madrid Spain
    • X. Yao, The University of Birmingham The Centre for Research in Computational Intelligence and Applications (CERCIA), School of Computer Science Birmingham UK

Evolutionary synthesis of low-sensitivity digital filters using adjacency matrix

Abstract  An evolutionary synthesis method to generate digital filters with low coefficient sensitivity is presented. The method uses
a chromosome coding based on the graph adjacency matrix representation. It is shown that the proposed chrom…

Abstract  An evolutionary synthesis method to generate digital filters with low coefficient sensitivity is presented. The method uses
a chromosome coding based on the graph adjacency matrix representation. It is shown that the proposed chromosome representation
enables to easily verify and avoid the generation of topologically invalid and non-computable individuals during the evolutionary
process. The efficiency of the proposed algorithm is tested in the synthesis of two low-pass digital filters and the results
are compared with other examples found in the literature.

  • Content Type Journal Article
  • DOI 10.1007/s12065-009-0028-x
  • Authors
    • Leonardo Bruno de Sá, Brazilian Army Technological Center Av das Américas, 28705, Guaratiba Rio de Janeiro 23020-470 Brazil
    • Antonio Mesquita, Federal University of Rio de Janeiro Rio de Janeiro Brazil

Rule base and adaptive fuzzy operators cooperative learning of Mamdani fuzzy systems with multi-objective genetic algorithms

Abstract  In this paper, we present an evolutionary multi-objective learning model achieving cooperation between the rule base and the
adaptive fuzzy operators of the inference system in order to obtain simpler, more compact and still accura…

Abstract  In this paper, we present an evolutionary multi-objective learning model achieving cooperation between the rule base and the
adaptive fuzzy operators of the inference system in order to obtain simpler, more compact and still accurate linguistic fuzzy
models by learning fuzzy inference adaptive operators together with rules. The multi-objective evolutionary algorithm proposed
generates a set of fuzzy rule based systems with different trade-offs between interpretability and accuracy, allowing the
designers to select the one that involves the most suitable balance for the desired application. We develop an experimental
study testing our approach with some variants on nine real-world regression datasets finding the advantages of cooperative
compared to sequential models, as well as multi-objective compared with single-objective models. The study is elaborated comparing
different approaches by applying non-parametric statistical tests for pair-wise. Results confirm the usefulness of the proposed
approach.

  • Content Type Journal Article
  • DOI 10.1007/s12065-009-0026-z
  • Authors
    • Antonio A. Márquez, University of Huelva Information Technologies Department Huelva Spain
    • Francisco A. Márquez, University of Huelva Information Technologies Department Huelva Spain
    • Antonio Peregrín, University of Huelva Information Technologies Department Huelva Spain

Extending a simple genetic cooperative-competitive learning fuzzy classifier to low quality datasets

Abstract  Exploiting the information in low quality datasets has been recently acknowledged as a new challenge in Genetic Fuzzy Systems.
Owing to this, in this paper we discuss the basic principles that govern the extension of a fuzzy rule b…

Abstract  Exploiting the information in low quality datasets has been recently acknowledged as a new challenge in Genetic Fuzzy Systems.
Owing to this, in this paper we discuss the basic principles that govern the extension of a fuzzy rule based classifier to
interval and fuzzy data. We have also applied these principles to the genetic learning of a simple cooperative-competitive
algorithm, that becomes the first example of a Genetic Fuzzy Classifier able to use low quality data. Additionally, we introduce
a benchmark, comprising some synthetic samples and two real-world problems that involve interval and fuzzy-valued data, that
can be used to assess future algorithms of the same kind.

  • Content Type Journal Article
  • DOI 10.1007/s12065-009-0024-1
  • Authors
    • Ana M. Palacios, Universidad de Oviedo Departamento de Informática 33071 Gijón Asturias Spain
    • Luciano Sánchez, Universidad de Oviedo Departamento de Informática 33071 Gijón Asturias Spain
    • Inés Couso, Universidad de Oviedo Departamento de Estadística e I.O. y D.M 33071 Gijón Asturias Spain

Evolution of visual controllers for obstacle avoidance in mobile robotics

Abstract  The purpose of this work is to automatically design vision algorithms for a mobile robot, adapted to its current visual context.
In this paper we address the particular task of obstacle avoidance using monocular vision. Starting fr…

Abstract  The purpose of this work is to automatically design vision algorithms for a mobile robot, adapted to its current visual context.
In this paper we address the particular task of obstacle avoidance using monocular vision. Starting from a set of primitives
composed of the different techniques found in the literature, we propose a generic structure to represent the algorithms,
using standard resolution video sequences as an input, and velocity commands to control a wheel robot as an output. Grammar
rules are then used to construct correct instances of algorithms, that are then evaluated using different protocols: evaluation
of trajectories performed in a goal reaching task, or imitation of a hand-guided trajectory. A genetic program is applied
to evolve populations of algorithms in order to optimize the performances of the controllers. The first results obtained in
a simulated environment show that the evolution produces algorithms that can be easily interpreted and which are clearly adapted
to the visual context. However, the resulting trajectories are often erratic, and the generalization capacities are poor.
To improve the results, we propose to use a two-phase evolution combining imitation and goal reaching evaluations, and to
add some constraints in the grammar rules to enforce a more generic behavior. The results obtained in simulation show that
the evolved algorithms are more efficient and more generic. Finally, we apply the imitation based evolution on real sequences
and test the evolved algorithms on a real robot. Though simplified by dropping the goal reaching constraint, the resulting
algorithms behave well in a corridor centering task, and show certain generalization capacities.

  • Content Type Journal Article
  • DOI 10.1007/s12065-009-0021-4
  • Authors
    • Renaud Barate, ENSTA 32 Bd Victor 75739 Paris Cedex 15 France
    • Antoine Manzanera, ENSTA 32 Bd Victor 75739 Paris Cedex 15 France