Now showing 1 - 10 of 35
  • Publication
    Zero is not a Four Letter Word: Studies in the Evolution of Language
    We examine a model genetic system that has features of both genetic programming and genetic regulatory networks, to show how various forms of degeneracy in the genotype-phenotype map can induce complex and subtle behaviour in the dynamics that lead to enhanced evolutionary robustness and can be fruitfully described in terms of an elementary algorithmic 'language'.
    Scopus© Citations 1  375
  • Publication
    Evolving behaviour trees for the Mario AI competition using grammatical evolution
    This paper investigates the applicability of Genetic Programming type systems to dynamic game environments. Grammatical Evolution was used to evolve Behaviour Trees, in order to create controllers for the Mario AI Benchmark. The results obtained reinforce the applicability of evolutionary programming systems to the development of artificial intelligence in games, and in dynamic systems in general, illustrating their viability as an alternative to more standard AI techniques.
      1521Scopus© Citations 59
  • Publication
    Introducing Semantic-Clustering Selection in Grammatical Evolution
    Semantics has gained much attention in the last few years and new advanced crossover and mutation operations have been created which use semantic information to improve the quality and generalisability of individuals in genetic programming. In this paper we present a new selection operator in grammatical evolution which uses semantic information of individuals instead of just the fitness value. The semantic traits of an individual are stored in a vector. An unsupervised learning technique is used to cluster individuals based on their semantic vector. Individuals are only allowed to reproduce with individuals from the same cluster to preserve semantic locality and intensify the search in a certain semantic area. At the same time, multiple semantic areas are covered by the search as there exist multiple clusters which cover different areas and therefore preserve semantic diversity. This new selection operator is tested on several symbolic regression benchmark problems and compared to grammatical evolution with tournament selection to analyse its performance.
      412Scopus© Citations 5
  • Publication
    Evolving Scale-Free Topologies using a Gene Regulatory Network Model
    A novel approach to generating scale-free network topologies is introduced, based on an existing artificial Gene Regulatory Network model. From this model, different interaction networks can be extracted, based on an activation threshold. By using an Evolutionary Computation approach, the model is allowed to evolve, in order to reach specific network statistical measures. The results obtained show that, when the model uses a duplication and divergence initialisation, such as seen in nature, the resulting regulation networks not only are closer in topology to scale-free networks, but also exhibit a much higher potential for evolution.
    Scopus© Citations 3  338
  • Publication
    Crossover, Population Dynamics and Convergence in the GAuGE System
    (Springer, 2004-06-30) ;
    This paper presents a study of the effectiveness of a recently presented crossover operator for the GAuGE system. This crossover, unlike the traditional crossover employed previously, preserves the association of positions and values which exists in GAuGE genotype strings, and as such is more adequate for problems where the meaning of an allele is dependent on its placement in the phenotype string. Results obtained show that the new operator improves the performance of the GAuGE system on simple binary problems, both when position-sensitive data is manipulated and not.
    Scopus© Citations 1  358
  • Publication
    Deep Evolution of Feature Representations for Handwritten Digit Recognition
    A training protocol for learning deep neural networks, called greedy layer-wise training, is applied to the evolution of a hierarchical, feed-forward Genetic Programming based system for feature construction and object recognition. Results on a popular handwritten digit recognition benchmark clearly demonstrate that two layers of feature transformations improves generalisation compared to a single layer. In addition, we show that the proposed system outperforms several standard Genetic Programming systems, which are based on hand-designed features, and use different program representations and fitness functions.
    Scopus© Citations 11  306
  • Publication
    Validation of a morphogenesis Model of Drosophila Early Development by a Multi-objective evolutionary Optimization Algorithm
    We apply evolutionary computation to calibrate the parameters of a morphogenesis model of Drosophila early development. The model aims to describe the establishment of the steady gradients of Bicoid and Caudal proteins along the antero-posterior axis of the embryo of Drosophila. The model equations consist of a system of non-linear parabolic partial differential equations with initial and zero flux boundary conditions. We compare the results of single- and multi-objective variants of the CMA-ES algorithm for the model the calibration with the experimental data. Whereas the multiobjective algorithm computes a full approximation of the Pareto front, repeated runs of the single-objective algorithm give solutions that dominate (in the Pareto sense) the results of the multi-objective approach. We retain as best solutions those found by the latter technique. From the biological point of view, all such solutions are all equally acceptable, and for our test cases, the relative error between the experimental data and validated model solutions on the Pareto front are in the range 3% − 6%. This technique is general and can be used as a generic tool for parameter calibration problems.
    Scopus© Citations 9  352
  • Publication
    Guidelines for defining benchmark problems in Genetic Programming
    The field of Genetic Programming has recently seen a surge of attention to the fact that benchmarking and comparison of approaches is often done in non-standard ways, using poorly designed comparison problems. We raise some issues concerning the design of benchmarks, within the domain of symbolic regression, through experimental evidence. A set of guidelines is provided, aiming towards careful definition and use of artificial functions as symbolic regression benchmarks.
      520Scopus© Citations 26
  • Publication
    Designing a Massive Dataset Framework for the Grammatical Evolution System
    (2016-07-06)
    In this study, a GE-framework is built, in an effort to apply it to huge datasets. Combining statistical techniques such as appropriate error measures and data splitting, population-based improvements such as mass parallelisation, and even specific techniques such as grammar design and repeat management, GE is applied for the first time to massive datasets, such as the Higgs dataset (eleven million samples).
      93
  • Publication
    Genetic Operators and Sequencing in the GAuGE System
    (IEEE, 2006-07-21) ;
    This paper investigates the effects of the mapping process employed by the GAuGE system on standard genetic operators. It is shown that the application of that mapping process transforms these operators into suitable sequencing searching tools. A practical application is analysed, and its results compared with a standard genetic algorithm, using the same operators. Results and analysis highlight the suitability of GAuGE and its operators, for this class of problems.
      387