Practical Smoothing: The Joys of P-splines

P-splines are a practical guide to smoothing, combining regression on B-splines with simple, discrete, roughness penalties. They were introduced in 1996 and have been used in many diverse applications. The authors demonstrate optimal smoothing using mixed model technology and Bayesian estimation, as well as classical tools like cross-validation and AIC. P-splines can also be used for regression on signals, varying-coefficient models, quantile and expectile smoothing, and composite links for grouped data. Penalties are the key to P-splines, and combining them with tensor products of B-splines extends these properties to multiple dimensions. An appendix offers a systematic comparison to other smoothers.

Format: Hardback
Length: 208 pages
Publication date: 18 March 2021
Publisher: Cambridge University Press

P-splines are a versatile and effective tool for smoothing data, combining regression on B-splines with simple, discrete, roughness penalties. Introduced in 1996 by the authors, they have found wide application in diverse fields. The regression basis simplifies handling non-normal data, making them suitable for generalized linear models. The authors demonstrate optimal smoothing using mixed model technology and Bayesian estimation, alongside classical techniques like cross-validation and AIC. They extend P-splines beyond simple smoothing to regression on signals, varying-coefficient models, quantile and expectile smoothing, and composite links for grouped data. Penalties are the core of P-splines, and with proper modifications, they can handle periodic and circular data, as well as shape constraints. Combining penalties with tensor products of B-splines further enhances their attractive properties in multiple dimensions. An appendix provides a systematic comparison to other smoothers.

P-splines are a versatile and effective tool for smoothing data, combining regression on B-splines with simple, discrete, roughness penalties. Introduced in 1996 by the authors, they have found wide application in diverse fields. The regression basis simplifies handling non-normal data, making them suitable for generalized linear models. The authors demonstrate optimal smoothing using mixed model technology and Bayesian estimation, alongside classical techniques like cross-validation and AIC. They extend P-splines beyond simple smoothing to regression on signals, varying-coefficient models, quantile and expectile smoothing, and composite links for grouped data. Penalties are the core of P-splines, and with proper modifications, they can handle periodic and circular data, as well as shape constraints. Combining penalties with tensor products of B-splines further enhances their attractive properties in multiple dimensions. An appendix provides a systematic comparison to other smoothers.

Weight: 452g
Dimension: 159 x 234 x 18 (mm)
ISBN-13: 9781108482950