The author has attempted to present a book that provides a non-technical introduction into the area of non-parametric density and regression function estimation. The application of these methods is discussed in terms of the S computing environment. Smoothing in high dimensions faces the problem of data sparseness. A principal feature of smoothing, the averaging of data points in a prescribed neighborhood, is not really practicable in dimensions greater than three if we have just one hundred data points. Additive models provide a way out of this dilemma; but, for their interactiveness and recursiveness, they require highly effective algorithms. For this purpose, the method of WARPing (Weighted Averaging using Rounded Points) is described in great detail.
- ISBN13 9781461287681
- Publish Date 19 October 2011 (first published 1 December 1990)
- Publish Status Active
- Publish Country US
- Imprint Springer-Verlag New York Inc.
- Edition Softcover reprint of the original 1st ed. 1991
- Format Paperback
- Pages 262
- Language English