商品詳細

Smoothing Techniques: With Implementation in S. Softcover reprint of the original 1st ed. 1991.

・ISBN 978-1-4612-8768-1 paper EUR 99.99

¥24,746.- (税込) *

著者・編者 Haerdle, Wolfgang,
シリーズ (Springer Series in Statistics)
出版社 (Springer-Verlag New York Inc., US)
出版年 2011
ページ数 262 pp.
ニュース番号 <M25-18161>

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.