Showing papers on "Statistical learning theory published in 1993"

Solvable Models of Artificial Neural Networks

Abstract: Solvable models of nonlinear learning machines are proposed, and learning in artificial neural networks is studied based on the theory of ordinary differential equations. A learning algorithm is constructed, by which the optimal parameter can be found without any recursive procedure. The solvable models enable us to analyze the reason why experimental results by the error backpropagation often contradict the statistical learning theory.

Application of Polynomial Neural Networks to Classification of Acoustic Warfare Signals

Abstract: : For both estimation and classification problems, the benefits of using artificial neural networks include inductive learning, rapid computation, and the ability to handle high-order and/or nonlinear processing. Neural networks reduce the need for simplifying assumptions that use a priori statistical models (such as 'additive Gaussian noise') or that neglect nonlinear terms, cross-coupling effects, and high-order dynamics. This report demonstrates the usefulness for acoustic warfare applications of an interdisciplinary approach that applies the rigorous theory and algorithms of statistical learning theory to the field of artificial neural networks. In particular, this approach provides two important results; (1) a generalized way of viewing neural modeling in terms of statistical function estimation, and (2) a constrained minimum- logistic-loss polynomial neural network (PNN) classification algorithm. These classification neural networks train rapidly, provide improved discrimination, and use an information-theoretic approach to limit structural complexity and thus avoid over-fitting training data. The report documents the successful application of these algorithms for the purpose of discriminating among broadband acoustic warfare signals and makes recommendations concerning further improvement of the algorithms. Artificial neural networks, Acoustic warfare, Sonar signal, Estimation, Machine learning, Processing, Classification, Modeling.