๐”– Bobbio Scriptorium
โœฆ   LIBER   โœฆ

Matrix logarithm parametrizations for neural network covariance models

โœ Scribed by Peter M Williams

Publisher
Elsevier Science
Year
1999
Tongue
English
Weight
141 KB
Volume
12
Category
Article
ISSN
0893-6080

No coin nor oath required. For personal study only.

โœฆ Synopsis

Neural networks are commonly used to model conditional probability distributions. The idea is to represent distributional parameters as functions of conditioning events, where the function is determined by the architecture and weights of the network. An issue to be resolved is the link between distributional parameters and network outputs. The latter are unconstrained real numbers whereas distributional parameters may be required to lie in proper subsets, or be mutually constrained, e.g. by the positive definiteness requirement for a covariance matrix. The paper explores the matrix-logarithm parametrization of covariance matrices for multivariate normal distributions. From a Bayesian point of view the choice of parametrization is linked to the choice of prior. This is treated by investigating the invariance of predictive distributions, for the chosen parametrization, with respect to an important class of priors.

๐Ÿ“œ SIMILAR VOLUMES

Bayesian Neural Network Models for Censo
Bayesian Neural Network Models for Censored Data
โœ David Faraggi; R. Simon; E. Yaskil; A. Kramar ๐Ÿ“‚ Article ๐Ÿ“… 1997 ๐Ÿ› John Wiley and Sons ๐ŸŒ English โš– 758 KB

Neural networks are considered by many to be very promising tools for classification and prediction. The flexibility of the neural network models often result in over-fit. Shrinking the parameters using a penalized likelihood is often used in order to overcome such over-fit. In this paper we extend

Fuzzy neural network models for liquefac
Fuzzy neural network models for liquefaction prediction
โœ M.S Rahman; Jun Wang ๐Ÿ“‚ Article ๐Ÿ“… 2002 ๐Ÿ› Elsevier Science ๐ŸŒ English โš– 243 KB

Integrated fuzzy neural network models are developed for the assessment of liquefaction potential of a site. The models are trained with large databases of liquefaction case histories. A two-stage training algorithm is used to develop a fuzzy neural network model. In the preliminary training stage,