Volker Tresp Siemens AG Central Research 81730 Munchen, Germany Volker. [email protected] We compare two regularization methods which can be used to improve the generalization capabilities of Gaussian mixture density estimates. The first method uses a Bayesian prior on the parameter space. We derive EM (Expectation Maximization) update rules which maximize the a posterior parameter probabili...

The purpose of this paper is to present a simple micromechanics-based model to estimate the effective thermal conductivity of macroscopically isotropic materials of matrix-inclusion type. The methodology is based on the well-established Mori-Tanaka method for composite media reinforced with ellipsoidal inclusions, extended to account for imperfect thermal contact at the matrix-inclusion interfa...

We demonstrate deterministic control over the photon-number distribution by interfering two coherent beams within a disordered photonic lattice. By sweeping a relative phase between two equal-amplitude coherent fields with Poissonian statistics that excite adjacent sites in a lattice endowed with disorder-immune chiral symmetry, we measure an output photon-number distribution that changes perio...

Based on an observation about the different effect of ensemble averaging on the bias and variance portions of the prediction error, we discuss training methodologies for ensembles of networks. We demonstrate the effect of variance reduction and present a method of extrapolation to the limit of an infinite ensemble. A significant reduction of variance is obtained by averaging just over initial c...

We present ensemble Bayesian model averaging (EBMA) and illustrate its ability to aid scholars in the social sciences to make more accurate forecasts of future events. In essence, EBMA improves prediction by pooling information from multiple forecast models to generate ensemble predictions similar to a weighted average of component forecasts. The weight assigned to each forecast is calibrated v...

A difference approximation that is second-order accurate in the time step h is derived for the general Ito stochastic differential equation. The difference equation has the form of a second-order random walk in which the random terms are non-linear combinations of Gaussian random variables. For a wide class of problems, the transition pdf is joint-normal to second order in h; the technique then...

Idealizing matter as a pressureless fluid and representing its motion by a peculiar-velocity field superimposed on a homogeneous and isotropic Hubble expansion, we apply (Lagrangian) spatial averaging on an arbitrary domain D to the (nonlinear) equations of Newtonian cosmology and derive an exact, general equation for the evolution of the (domain dependent) scale factor aD (t). We consider the ...