By Topic

IJCNN-91-Seattle International Joint Conference on Neural Networks

8-12 July 1991

Go

Filter Results

Displaying Results 1 - 25 of 383
  • Learning to generate subgoals for action sequences

    Publication Year: 1991
    Cited by:  Papers (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (27 KB)

    Summary form only given. None of the existing learning algorithms for neural networks in time-varying environments addresses the problems of learning to 'divide and conquer'. Algorithms based on pure gradient descent or on adaptive critic methods are not suitable for dynamic control problems with long time lags between actions and consequences, and that there is a need for algorithms that perform ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A neural network for interpretation of multi-meaning words

    Publication Year: 1991, Page(s):491 - 496 vol.2
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (373 KB)

    A neural network is proposed that can interpret words with multiple meanings correctly by using contextual information. The network learns to recognize its present context and utilizes the contextual information to interpret the words correctly. Representation of the context is achieved by making a unit respond to a series of successively presented words. Interpretation of an input word is achieve... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Unsupervised learning for the visual cortex (layer IV): model and simulations

    Publication Year: 1991, Page(s):613 - 618 vol.2
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (415 KB)

    The authors present a statistical model to study the evolution of orientation columns in layer IV of the visual cortex after birth by simulating, on the retina, the first visual experiences of an animal (darkness, biased, and normal visual environments). The results are consistent with biological experiments. The authors show the necessity of introducing a new nonlinear model in which both vertica... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Computational model of a biologically plausible cognitive map

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (48 KB)

    Summary form only given, as follows. Humans develop spatial cognitive maps using knowledge gathered while repeatedly observing their environment. This knowledge may come from landmarks present in the environment, the relative position(s) of these landmarks, or available paths along which an observer can reach these landmarks. Recent neurophysiological findings in rats suggest that there are specia... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Neural network routing for multiple stage interconnection networks

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (48 KB)

    Summary form only given, as follows. A Hopfield model neural network can be useful as a form of parallel computer. Such a neural network may be capable of arriving at a problem solution which much more speed than conventional, sequential approaches. This concept has been applied to the problem of generating control bits for a multistage interconnection network. A Hopfield model neural network has ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Neural network analysis of DNA sequences

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (45 KB)

    Summary form only given. Attention is given to a project which investigates the applicability of recently developed neural network programs to analyze DNA sequences in a few simple cases to determine parameters such as learning times. Specifically, the ability of various macromolecules to bind to nucleic acid with a recognizable pattern has been considered. The group of enzymes known as restrictio... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A neural network approach to the vehicle dispatching problem

    Publication Year: 1991
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (45 KB)

    Summary form only given. Decision making with respect to the dispatching of vehicles and crews is still mainly in the realm of human expertise. On the basis of their own experience, the authors suggest that it seems very difficult to explicitly model that expertise via a symbolic approach. They thus propose an alternative neural network model as a subsymbolic and empirical alternative for modeling... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A comparison of MLP and ID3-derived approaches for ECG classification

    Publication Year: 1991
    Cited by:  Papers (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (36 KB)

    Summary form only given. A preprocessing technique was used to deal with time-shift invariance for an MLP (multilayer perceptron), and its classification performance was compared to that of C4 (an ID3-derived approach). The comparison is in terms of generalization capability and sensitivity to noise, and experimental results show a superior performance of MLP techniques for ECG (electrocardiograph... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • JackKnife method for validating neural network models

    Publication Year: 1991
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (36 KB)

    Summary form only given. Most methods for validating neural networks rely on the exclusion of a portion of the data throughout a portion or all of the network training process. This approach can be somewhat wasteful, particularly when the cost of samples can exceed a million dollars each. It is suggested that a more appropriate (and efficient) validation method is the JackKnife method. Although th... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Applications of neural networks to F-16 avionics

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (57 KB)

    Summary form only given. The author notes that the software support division of the Ogden Air Logistics Center has developed a series of neural network applications for which traditional approaches have failed. In addition, this division has developed a methodology for identifying such problems, and a generic approach for solving them.<> View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Cooperative schemes for conventional and neural expert systems

    Publication Year: 1991
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (57 KB)

    Summary form only given. Cooperative schemes between neural net models and fuzzy expert systems were investigated. Performance of a specific neural net model with back-propagation models is discussed. Information is initially received by the expert system (ES) in a form that is useful for certain tasks. Information is represented in more than one form, reflecting interpretations of uncertainties a... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Approximate representation of a continuous function by a neural network with scaled or unscaled sigmoid units

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (50 KB)

    Summary form only given. The author investigates the capability of three-layered neural networks with a scaled or unscaled sigmoid activation function of the hidden layer units in uniformly approximate representation of continuous functions. For the approximation on a compact set, any sigmoid function can be the activation function without scaling. Even for the approximation of continuous function... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A design principle of the analog neural network for combinatorial optimization: findings from a study on the placement problem

    Publication Year: 1991
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (50 KB)

    Summary form only given. A design principle of a Hopfield model to be used for combinatorial optimization is discussed through its application to a circuit block placement problem. The authors propose several means to improve the energy function of the Hopfield model, and a tuning rule of its parameters. Computer simulations show that the Hopfield model yields quite satisfactory solutions to the h... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Neural network learning of structure-activity relationships

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (50 KB)

    Summary form only given, as follows. The author describes a pattern recognition problem using counter-propagation feature extraction networks applied to toxicological data.<> View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An application of Hopfield network to discrete Fourier transform

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (40 KB)

    Summary form only given. An application of a Hopfield network to the discrete Fourier transform was studied. First, the commonalities and differences were defined by a study and comparison of the discrete Fourier transform and Hopfield network structures, and then, in order to investigate the difference, a numerical analysis was conducted. As a result, it became clear that since the coefficient ma... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Polynomial neural networks for signal processing in chaotic backgrounds

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (40 KB)

    Summary form only given. Neural network signal processors can be trained to detect threshold signals in chaotic backgrounds where accurate global prediction, based upon learned behavior, may be possible. The neural network signal processing method is based upon the synthesis of a polynomial neural network (PNN) for global prediction of the chaotic background which is then subtracted in order to en... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Map learning using associative memory neural network

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (41 KB)

    Summary form only given. Map learning using associative memory is considered. Given a source location and a destination location to be visited and its associated visiting path, an associative memory neural network which can remember and recall all possible paired-location combinations is constructed. Kth nearest neighbor transformation (Knn) is used to transfer the input paired locations to a vect... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A novel IDMLP model and its application to the recognition of Chinese character

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (41 KB)

    Summary form only given. A learning method for the recognition of printed Chinese characters using an input-driven multilayer perceptron (IDMLP) model has been proposed, and the circuit representing the learning result has been designed. In learning, the extracted features from Chinese characters are used as inputs, and the synapse weight is an integer value. It is possible to implement the learni... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The effect of ANN hidden layer neurons on boundary identification for constrained optimization

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (40 KB)

    Summary form only given. The process of constrained optimization involves first finding a feasible region which satisfies all design criteria and then optimizing the design goal while remaining in this feasible region. The constraint boundaries, which are the borders between feasible and infeasible regions, were identified using the pattern classification capability of backpropagation networks. In... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Gestures and neural networks in human-computer interaction

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (40 KB)

    Summary form only given. Neural networks are recognized as being able to learn to solve classification problems, and their generalization properties make them suitable for interpreting imprecise input values. These features of neural networks were utilized by applying networks to the problem of recognizing gestural input. The signs made by a user are interpreted and classified by the network, allo... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A hierarchical interaction architecture for pattern recognition

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (54 KB)

    Summary form only given. Proposes a hierarchical interaction neural network (HINT) model for complex pattern recognition. HINT recognizes handwritten Hangul characters in a context-dependent way. HINT consists of two subnets, FD-net and HC-net, which detect Hangul alphabets and implement Hangul character composition rules, respectively. The underlying theme is that complex pattern recognition invo... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A plant modeling with a fuzzy associative memory system

    Publication Year: 1991
    Cited by:  Papers (1)  |  Patents (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (54 KB)

    Summary form only given. An attempt was made to demonstrate a useful qualitative model using causal concepts for modeling the operations of a physical plant. In order to control and monitor plants, support based on the intuitive, qualitative perspective of causal relations becomes important, along with the usual numerical, logical approach. This goal is realized by using a qualitative model, summa... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Correct reaction neural network and its implementation

    Publication Year: 1991
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (44 KB)

    Summary form only given. A novel method has been devised to analyze and to design Hopfield-like neural networks in which globally optimal solutions to specific problems are guaranteed. As the first part of a project to construct practical circuits for massively parallel analog computation, chips containing a 4-bit A/D converter have been constructed for demonstration purposes.<> View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Implementation of the relaxation process by the interactive activation and competition network

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (44 KB)

    Summary form only given. There exist certain common properties in the relaxation process and the neural network technique. A mapping method that makes the interactive activation and competition network perform the relaxation process has been proposed. With this method, neural network technology can be easily adapted to solve the many problems which have already been solved by the relaxation proces... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Neurocomputing and computer-based tutoring: modeling learner performance

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (54 KB)

    Summary form only given. Preliminary research results have been obtained which indicate that neural networks represent a promising approach for use in intelligent tutoring systems (ITS). A neural network was used to analyze learner responses to subject matter posed by a second language tutor and to classify learners into different language competence and motivational categories. An ITS could use t... View full abstract»

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