By Topic

Evolutionary Computation, 1995., IEEE International Conference on

Date Nov. 29 1995-Dec. 1 1995

Go

Filter Results

Displaying Results 1 - 25 of 63
  • Index [of authors]

    Publication Year: 1995
    Save to Project icon | Request Permissions | PDF file iconPDF (179 KB)  
    Freely Available from IEEE
  • Application of genetic algorithms to system identification

    Publication Year: 1995 , Page(s): 777 - 782 vol.2
    Cited by:  Papers (4)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (396 KB)  

    System identification is a pre-requisite to the analysis of a dynamic system and design of an appropriate controller for improving its performance. In conventional identification methods, a model structure is selected and the parameters of that model are calculated by optimising an objective function. This process usually requires a large set of input/output data from the system which is not always available. In addition the obtained parameters may be only locally optimal. In this work genetic algorithms are applied to system identification. A system is assumed to have an ARMAX model, the parameters of which are obtained using the search process of the genetic algorithms. The method developed is presented and results of its application to a number of experimental systems are described. The results obtained are quite encouraging View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Evolving neural network controllers

    Publication Year: 1995 , Page(s): 579 - 583 vol.2
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (476 KB)  

    An emerging design paradigm uses evolutionary processes to search for optima in design space. The evolutionary technique has the advantage of being a declarative paradigm; the user specifies the task, and a genetic algorithm searches for an optimum solution. Normal techniques require the definition of the controller, and this is computationally expensive. We use a genetic algorithm to design a neural network-based controller for a hexapod robot. The robot must perform the task of moving from a start position to a goal position, under varying degrees of simulated instrument and sensor noise. The findings show that it is possible to embed a degree of noise tolerance into the solution. This is useful in situations where the environment of the robot may change over time View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Gait coordination of hexapod walking robots using mutual-coupled immune networks

    Publication Year: 1995 , Page(s): 672 - 677 vol.2
    Cited by:  Papers (3)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (588 KB)  

    Biological information processing systems are amongst the ultimate decentralized systems, and are expected to provide various fruitful ideas for engineering fields, especially robotics. Among these systems, the brain-nervous system and genetic system have already been widely used by modeling as neural networks and genetic algorithms, respectively. The immune system also plays an important role in coping with a dynamically changing environment by constructing self-nonself recognition networks among different species of antibodies. This system also has a lot of interesting features such as learning, self-organizing abilities and so on viewed from the engineering standpoint. However, the immune system has not yet been applied to engineering fields. We propose a new hypothesis concerning the structure of the immune system, called the mutual-coupled immune network hypothesis, based on recent studies on immunology. We apply this idea to gait acquisition of a hexapod walking robot as a practical example. Finally, the feasibility of our proposed method is confirmed by simulations View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Genetic operators using viral models

    Publication Year: 1995 , Page(s): 652 - 656 vol.2
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (464 KB)  

    The use of inversion operators based loosely on viral models are compared using the travelling salesman problem. Using statistical methodologies, some of the operators are found to be preferable in that they offered, on average, significantly faster convergence combined with less likelihood to “plateau” early View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Genetic algorithms incorporating a pseudo-subspace method

    Publication Year: 1995 , Page(s): 557 - 560 vol.2
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (412 KB)  

    GA performance in high-dimensional optimisation problems can be enhanced by the use of a `pseudo subspace' technique. The method works by projecting the parameter space onto a lower dimensional subspace in the first stages of the optimisation process, in order to allow the GA search to discover the most promising area of the solution space. Subsequently, the dimensionality of the model is progressively increased until a predetermined limit is reached. Comparison between the pseudo-subspace procedure and a conventional GA, using two different GA implementations, shows the former to be more successful when applied to two geophysical problems characterised by different solution-space geometry and mathematics. This technique could be easily transferred to different image processing or pattern recognition problems where geometrical relationships between the parameters are maintained View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Phenotypic forking genetic algorithm (p-fGA)

    Publication Year: 1995 , Page(s): 566 - 572 vol.2
    Cited by:  Papers (2)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (572 KB)  

    Proposes a new type of multi-population genetic algorithm, the p-fGA (phenotypic forking GA), an extension of the previously proposed g-fGA (genotype forking GA). Both the g-fGA and the p-fGA are designed to solve multi-modal problems which are difficult to solve by traditional GAs. We use multi-population schemes that include one parent population with a blocking mode and one or more child populations with a shrinking mode. The g-fGA defines its sub-space for each population by a “salient schema” within the genotypic search space. In contrast to this, the p-fGA defines its sub-space by a “neighborhood hypercube” around the current best individual in the phenotypic search space. Empirical results show that the p-fGA has a fairly good performance, as does the g-fGA, and the variable-resolution p-fGA has the capability of searching with high resolution and can improve on the local search capability in a genetic search View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Applying logic grammars to induce sub-functions in genetic programming

    Publication Year: 1995 , Page(s): 737 - 740 vol.2
    Cited by:  Papers (5)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (424 KB)  

    Genetic programming (GP) is a method of automatically inducing S-expressions in LISP to perform specified tasks. The problem of inducing programs can be reformulated as a search for a highly fit program in the space of all possible programs. This paper presents a framework in which the search space can be specified declaratively by a user. Its application in inducing sub-functions is detailed. The framework is based on a formalism of logic grammars and it is implemented as a system called LOGENPRO (LOgic grammar-based GENetic PROgramming system). The formalism is powerful enough to represent context-sensitive information and domain-dependent knowledge. This knowledge can be used to accelerate the learning speed and/or improve the quality of the programs induced. The system is also very flexible and programs in various programming languages can be acquired. Automatic discovery of sub-functions is one of the most important research areas in GP. An experiment is used to demonstrate that LOGENPRO can emulate Koza's (1992, 1994) automatically defined functions (ADF). Moreover, LOGENPRO can employ knowledge such as argument types in a unified framework. An experiment shows that LOGENPRO has a superior performance to that of ADF when more domain-dependent knowledge is available View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Discovery of self-replicating structures using a genetic algorithm

    Publication Year: 1995 , Page(s): 678 - 683 vol.2
    Cited by:  Papers (6)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (452 KB)  

    Previous computational models of self-replication in cellular spaces have been manually designed, a very difficult and time-consuming process. This paper introduces the use of genetic algorithms to discover automata rules that govern emergent self-replicating processes. Given dynamically evolving automata, identification of effective fitness functions for self-replicating structures is a difficult task, and we give one solution to this problem. A model consisting of movable automata embedded in a cellular space is introduced and discussed in this context. A genetic algorithm using two fitness criteria was applied to automate rule discovery. After parameter tuning, 6 self-replicating structures consisting of 2, 3 and 4 automata were discovered over a course of 75 genetic algorithm runs. These results indicate that the fitness functions employed are effective and that genetic algorithms can be used to successfully discover rules for self-replicating structures View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Machine requirements planning and workload assignment using genetic algorithms

    Publication Year: 1995 , Page(s): 711 - 715 vol.2
    Cited by:  Patents (4)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (356 KB)  

    This paper presents a genetic approach to determining the optimal number of machines required in a manufacturing system for meeting a specified production schedule. This use of genetic algorithms is illustrated by solving a typical machine requirements planning problem. Comparison of the respective results obtained by using the proposed approach and a standard mixed-integer programming package shows that the proposed approach is indeed an effective means for optimal manufacturing systems design View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Competition and mutualism in a simulation of adaptive artificial organisms

    Publication Year: 1995 , Page(s): 695 - 700 vol.2
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (512 KB)  

    In the field of artificial life, evolutionary programming like genetic programming is a well known approach to the “central place food foraging” problem of artificial organisms. In most of these approaches, the organisms follow the same strategy. In this paper, an artificial organism is modeled with a classifier system, and the process of acquiring cooperative behavior is simulated by putting different kinds of organisms in an area. In the experiments, it is observed that one type of organism cooperates with another type of organism to protect their lives View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Self-adaptive genetic algorithm learning in game playing

    Publication Year: 1995 , Page(s): 814 - 818 vol.2
    Cited by:  Papers (4)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (484 KB)  

    Genetic algorithms (GAs) are known to be effective search methods that are also robust and efficient. We introduce a self-adaptive function for conventional GAs. A dynamic fitness technique helpful for continuous evolution and robust solution is also presented. We expect to improve the quality of GA searches in solving direct competitive problems. We tested our idea by using it to play the game Othello, a typical problem with the direct competitive properties. Experimental results show that our method is better than traditional approaches View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A fitness scaling method based on a span measure

    Publication Year: 1995 , Page(s): 561 - 565 vol.2
    Cited by:  Papers (2)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (324 KB)  

    This paper describes a new fitness scaling method based on a characteristic of a population of chromosomes known as the span. This transformation is both scale and translation invariant. Its behavior is illustrated on two complex multimodal functions and a comparison is provided with a power law scaling method View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Hybridized crossover-based search techniques for program discovery

    Publication Year: 1995 , Page(s): 573 - 578 vol.2
    Cited by:  Papers (5)  |  Patents (2)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (520 KB)  

    Addresses the problem of program discovery as defined by genetic programming. By combining a hierarchical crossover operator with two traditional single-point search algorithms (simulated annealing and stochastic iterated hill climbing), we have solved some problems by processing fewer candidate solutions and with a greater probability of success than genetic programming. We have also enhanced genetic programming by hybridizing it with the simple idea of hill climbing from a few individuals, at a fixed interval of generations View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Solving vehicle routing problems with genetic algorithms

    Publication Year: 1995 , Page(s): 788 - 793 vol.2
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (352 KB)  

    Many transportation problems, such as the travelling salesman problem, are computationally hard but often solvable quickly, although with less certainty, by heuristic methods. Genetic algorithms fall into this category and generate results with favourable scaling behaviour. We apply a two-level genetic algorithm to an advanced transportation problem, an example of the General Pickup and Delivery Problem. We discuss the formulation of the problem as an evolutionary one, show that the results scale well with size and that application to real-world situations is within reach View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Genetic logic programming

    Publication Year: 1995 , Page(s): 728 - 732 vol.2
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (440 KB)  

    Genetic logic programming (GLP) is a new method which applies the genetic algorithm paradigm to declarative programming-specifically to evolve populations of Prolog programs. This paper examines GLP applied to natural language understanding to illustrate the power, issues and limitations of GLP. Populations of Prolog query interpreters evolve to respond more correctly to queries about Aesop's fable “The Fox and the Crow”. The interpreters process parsed text and consult a general knowledge-base. The gene pool consists of a large set of Prolog rules and facts which are tentatively proposed as being `useful' for interpretation. Essentially, interpreters act as an interface between queries, knowledge bases and the text. Closure and termination are addressed at the level of design of the gene pool, and various Prolog options. Fitness amounts to a score on a high-school-like “comprehension test”, with special care needed to deal with redundant and dependent answers, and with an eye to rewarding correct higher-level abstractions View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Polycell placement for analog LSI chip designs by genetic algorithms and tabu search

    Publication Year: 1995 , Page(s): 716 - 721 vol.2
    Cited by:  Papers (4)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (488 KB)  

    Presents a genetic approach to the placement problem of analog LSI chip designs. The emphasis is the application of our approach to a real-world problem. The objective of this problem is to minimize the overall net length and the layout area under many electrical constraints among elements, and under the condition of making the subsequent routing task easy. Our approach is based on a polycell placement style to make the routing easy and it minimizes several cost functions. Computational experiments show the efficiency of our genetic algorithm (GA) compared to the tabu search (TS). The concatenation of GA and TS, and a dynamic adjustment of the mutation rate are introduced, and their effectiveness is shown View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • On the application of genetic programming to chemical process systems

    Publication Year: 1995 , Page(s): 701 - 706 vol.2
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (432 KB)  

    A genetic programming approach is utilised to develop mathematical models of chemical process systems. Having discussed genetic programming in general, two examples are used to reveal the utility of the technique. It is shown how the method can discriminate between relevant and irrelevant process inputs, evolving to yield parsimonious model structures that accurately represent process characteristics. This removes the need for restrictive assumptions about the form of the data and the structure of the required model. In addition, as the technique determines complex nonlinear relationships in the data, non-intuitive process features are revealed with comparative ease View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Hierarchical description of two dimensional shapes using a genetic algorithm

    Publication Year: 1995 , Page(s): 637 - 640 vol.2
    Cited by:  Papers (3)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (312 KB)  

    A description method for arbitrary two dimensional shapes is proposed in this paper. When a 2D shape is given as a silhouette, its structure is automatically approximated by the use of a set of rectangles by the proposed method. Sizes, positions and rotational angles of the rectangles which approximate adequately the given 2D shape are searched by a genetic algorithm; GA. In our coding of GA, a chromosome of each individual is a bit string corresponding to parameter sets of several rectangles. Through a generation iteration, accuracy of approximation of the given 2D shape is improved. The total number of rectangles to be used for description is assumed to be given before shape description. By changing the total number of rectangles, hierarchical description of given 2D shapes is achieved. This method can be applied to shape description and object recognition in the field of computer vision and to abstraction of 2D shapes in the field of artistic applications by the use of computers View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Learning the parameters for a gradient-based approach to image segmentation from the results of a region growing approach using cultural algorithms

    Publication Year: 1995 , Page(s): 819 - 824 vol.2
    Cited by:  Papers (3)  |  Patents (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (472 KB)  

    There are two basic approaches to image segmentation, region based and neighborhood based. Region based approaches require less a priori knowledge about the scene than neighborhood based approaches but are computationally more expensive. In cases where there is little prior knowledge about properties of the image, one is often forced to use region growing approaches. We use cultural algorithms, a form of evolutionary computation based upon principles of cultural evolution, as the basis for learning the parameters for a neighborhood based approach to image segmentation from the results of a region growing approach. Specifically, parameters for a differential gradient method utilizing the Sobel operator are learned from the results of a region growing approach. The prototype is applied to a sequence of real world images, taken from archaeological excavations of a prehistoric site in order to extract spatial activity areas in the site. A region growing approach is applied first to the images, and then a cultural algorithm is used to extract the parameters for use by a gradient method for those images. The resulting performance of the gradient method produced a correspondence of over 95% with that of the original View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Evolving complex neural networks that age

    Publication Year: 1995 , Page(s): 590 - 595 vol.2
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (468 KB)  

    The combination of the broad problem-searching capabilities of a genetic algorithm with the local maxima location capabilities of a hill-climbing algorithm can be a powerful technique for solving classification problems. Producing a number of specialist artificial neural networks, each an expert on one category, can be beneficial when solving problems in which the categories are distinct. This paper describes combining genetic algorithms, hill climbing and sets of specialist artificial neural networks to solve a difficult character recognition problem. It also describes a method by which the effects of a large “elite” sub-population can be counter-balanced by using an aging coefficient in the fitness calculation View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Using genetic programming to evolve board evaluation functions

    Publication Year: 1995 , Page(s): 747 - 752 vol.2
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (464 KB)  

    Employs the genetic programming paradigm to enable a computer to learn to play strategies for the ancient Egyptian boardgame Senet by evolving board evaluation functions. Formulating the problem in terms of board evaluation functions made it feasible to evaluate the fitness of game playing strategies by using tournament-style fitness evaluation. The game has elements of both strategy and chance. Our approach learns strategies which enable the computer to play consistently at a reasonably skillful level View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An evolutionary algorithm for function inversion and boundary marking

    Publication Year: 1995 , Page(s): 794 - 797 vol.2
    Cited by:  Papers (7)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (320 KB)  

    We present an evolutionary algorithm for distributing points evenly on a surface f(x)=c. Applications include function inversion (where system inputs are sought which produce a desired output) and boundary detection for e.g. control systems or power-system security assessment, in which boundaries of a safe operating region are to be avoided View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An evolutionary and cooperative agents model for optimization

    Publication Year: 1995 , Page(s): 668 - 671 vol.2
    Cited by:  Papers (2)  |  Patents (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (232 KB)  

    The authors propose the use of genetic algorithms (GA) to optimize another algorithm for optimization. The aim is to integrate the approach introduced by Dorigo et al., known as the ant system, with GA, exploiting the cooperative effect of the latter and the evolutionary effect of GA. An ant algorithm aims to solve problems of combinatorial optimization by means of a population of agents/processors that work parallel without a supervisor in a cooperative manner. A genetic algorithm aims to optimize the performance of the ant population by selecting optimal values for its parameters by means of evolution of the genetic patrimony associated with each single agent. The approach has been applied to the traveling salesman problem; results and comparisons with the original method are presented View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An expansion operator for interactive evolution

    Publication Year: 1995 , Page(s): 798 - 802 vol.2
    Cited by:  Papers (4)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (560 KB)  

    We demonstrate how interactive evolution can be applied to the extrapolation and growth of graphical models. In addition to the mutation and recombination operator for interactive evolution, we introduce a new operator termed expansion and show it to play a significant role in interactive evolution. The expansion operator predicts new future models on the basis of time series analyses of evolutionary processes. We demonstrate the application of this operator in the context of interactive evolution to predict or extrapolate new graphical models. The primary value of this operator is that it provides the user with a tool for exploring a design space that employs the general evolutionary constraint of incremental change according to a fitness value. Such an operator might, therefore, also be applicable to research on environmental change, biological development and growth, and engineering construction projects. The utility of the expansion operator suggests that it will open up new areas for further research in the evolutionary variation process View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.