Buy a book... 
| Bayesian Nonparametrics via Neural Networks | Buy a book...
|
| Author(s): | H. K. H. Lee |
| Publisher: | SIAM |
| ISBN: | 0898715636 |
| Format: | softback |
| 96pp | |
| Price: | $42.50 |
| Review Date: | 16 July 2004 |
| Review: | Herbert Lee approaches neural networks from a statistical perspective, treating them with the rigor that is normally accorded to statistical models. One of the themes is that a neural network can be viewed as a finite-dimensional approximation to an infinite-dimensional model that can be useful for solving problems in regression and classification. The first section introduces neural networks within the statistical context of nonparametric regression and classification. The remainder examines important statistical modelling issues for Bayesian neural networks and, in particular, the choice of prior and the choice of model. The book is for practising statisticians and researchers, computational scientists and data moniers. |