Partial differential equations (PDEs) with Dirichlet boundary conditions defined on boundaries with simple geomerty have been succesfuly treated using sigmoidal multilayer perceptrons in previous works [1, 2]. This article deals with the case of complex boundary geometry, where the boundary is determined by a number of points that belong to it and are closely located, so as to offer a reasonabl...

The determination of the initial weights is an important issue in multilayer perceptron design. Recently, we have proposed a new approach to weight initialization based on discriminant analysis techniques. In this paper, the performances of multilayer perceptrons (MLPs) initialized by non-parametric discriminant analysis are compared to those of randomly initialized MLPs using several synthetic...

In this paper, we introduce several hybrid connectionist-structural acoustic models for context-independent phone-like units. The structural part has been modeled with Markov chains or nite state networks learned by grammatical inference techniques. A multilayer perceptron or a committee of multilayer perceptrons is used to estimate the emission probabilities of the structural models. We compar...

IEEE TNN A172Rev K Nearest Neighbours Directed Noise Injection in Multilayer Perceptron Training M. Skurichina1, .Raudys2 and R.P.W. Duin1 1Pattern Recognition Group, Department of Applied Physics, Delft University of Technology, P.O. Box 5046, 2600GA Delft, The Netherlands. E-mail: [email protected], [email protected] 2Department of Data Analysis, Institute of Mathematics and Informa...

Visualization of high-dimensional data is a major task in data mining. The main idea of visualization is to map data from the highdimensional space onto a certain position in a low-dimensional space. From all mappings, only those that lead to maps that are good approximations of the data distribution observed in the high-dimensional space are of interest. Here, we present a mapping scheme based...

We propose a general class of multilayer feed-forward neural networks where the combination of inputs in every node is formed by hybrid linear and nonlinear (of the morphological/rank type) operations. We demonstrate that this structure ooers eecient solutions to pattern classiication problems by requiring fewer nodes or fewer parameters to estimate than those needed by multilayer perceptrons. ...

Partial differential equations (PDEs) with boundary conditions (Dirichlet or Neumann) defined on boundaries with simple geometry have been successfully treated using sigmoidal multilayer perceptrons in previous works. This article deals with the case of complex boundary geometry, where the boundary is determined by a number of points that belong to it and are closely located, so as to offer a r...

We study the dynamics of on-line learning for a large class of neural networks and learning rules, including backpropagation for multilayer perceptrons. In this paper, we focus on the case where successive examples are dependent, and we analyze how these dependencies affect the learning process. We define the representation error and the prediction error. The representation error measures how w...

Previous work on analog VLSI implementation of multilayer perceptrons with on-chip learning has mainly targeted the implementation of algorithms such as back-propagation. Although back-propagation is efficient, its implementation in analog VLSI requires excessive computational hardware. It is shown that using gradient descent with direct approximation of the gradient instead of back-propagation...