Smoothing Splines: Methods and Applications

ISBN-13: 9781420077551 / Angielski / Twarda / 2010 / 384 str.

Yuedong Wang

Opis

Driven by many sophisticated applications and fuelled by modern computing power, many powerful and flexible nonparametric and semiparametric modeling techniques have been developed to relax traditional parametric models and exploit possible hidden structure. As one of the most popular nonparametric techniques, the smoothing spline model has attracted a great deal of research attention in recent years. This book covers basic nonparametric spline models, such as polynomial spline, periodic spline, thin-plate spline, L- spline, partial spline, and smoothing spline ANOVA for both Gaussian data and data from exponential families. It also presents smoothing spline models with unequal variance and correlated random errors as well as advanced models, including semiparametric linear and nonlinear mixed-effects models and nonparametric and semiparametric nonlinear regression models. In addition, the author describes vector spline models for multivariate observations and functional linear models.

Spis treści

Introduction. Smoothing Spline Regression. Smoothing Parameter Selection and Inference. Smoothing Spline ANOVA. Spline Smoothing with Heteroscedastic and/or Correlated Errors. Generalized Smoothing Spline ANOVA. Smoothing Spline Nonlinear Regression. Semiparametric Regression. Semiparametric Mixed-Effects Models. Appendices. References. Indices.

O autorze

Yuedong Wang is a professor and the chair of the Department of Statistics and Applied Probability at the University of California–Santa Barbara. Dr. Wang is an elected fellow of the ASA and ISI, a fellow of the RSS, and a member of IMS, IBS, and ICSA. His research covers the development of statistical methodology and its applications.