Bayesian Nonparametrics via Neural Networks. Herbert K. H. LEE, Philadelphia: American Statistical Association and Society for Industrial and Applied Engineering, 2004, ISBN 0-89871-563-6, x + 96 pp., $42.50 ($29.75 for SIAM members).
Herbert K. H. Lee's monograph is a crisp introduction to the inference of neural networks using the Bayesian framework. The book can be broadly divided into an introductory section, comprising Chapters 1 and 2; a section describing the main material, comprising Chapter 3 on the problem of inferring a model from the data, and Chapter 4 on the problem of selecting and combining models; and a concluding section, comprising of Chapter 5 and the ...