Helsinki University of Technology

About: Helsinki University of Technology is a based out in . It is known for research contribution in the topics: Artificial neural network & Finite element method. The organization has 8962 authors who have published 20136 publications receiving 723787 citations. The organization is also known as: TKK & Teknillinen korkeakoulu.

Well-conditioned Muller formulation for electromagnetic scattering by dielectric objects

Abstract: Numerical solution of electromagnetic scattering by homogeneous dielectric objects with the method of moments (MoM) and Rao-Wilton-Glisson (RWG) basis functions is discussed. It is shown that the low-frequency breakdown associated to the MoM solution of scattering by dielectric objects can be avoided by the classical Muller formulation without the loop-tree or loop-star basis functions. Two variations of the Muller formulation, T-Mu/spl uml/ller and N-Muller, are considered. It is demonstrated that only the N-Muller formulation with the Galerkin method and RWG functions gives stable solution. Discretization of the N-Muller formulation leads to a well-conditioned matrix equation and rapidly converging iterative solutions on a wide frequency range from very low frequencies to microwave frequencies. At zero frequency, the N-Muller formulation decouples into the electrostatic and magnetostatic integral equations.

General-purpose computation with neural networks: a survey of complexity theoretic results

Abstract: We survey and summarize the literature on the computational aspects of neural network models by presenting a detailed taxonomy of the various models according to their complexity theoretic characteristics. The criteria of classification include the architecture of the network (feedforward versus recurrent), time model (discrete versus continuous), state type (binary versus analog), weight constraints (symmetric versus asymmetric), network size (finite nets versus infinite families), and computation type (deterministic versus probabilistic), among others. The underlying results concerning the computational power and complexity issues of perceptron, radial basis function, winner-take-all, and spiking neural networks are briefly surveyed, with pointers to the relevant literature. In our survey, we focus mainly on the digital computation whose inputs and outputs are binary in nature, although their values are quite often encoded as analog neuron states. We omit the important learning issues.