1: \begin{center}

2: {\large\bf

3: An Introduction to (Smoothing Spline) ANOVA Models in RKHS, With 

4: Examples in Geographical Data, Medicine, Atmospheric Science 

5: and Machine Learning}.

6: \end{center}

7: {\large\bf

8: \begin{center}

9: Grace Wahba \\

10: Department of Statistics \\

11: University of Wisconsin-Madison \\

12: {\tt http://www.stat.wisc.edu/\~{}wahba}

13: \end{center}

14: }

15: \begin{center}

16: This note has appeared in the Proceedings of the 13th 

17: IFAC Symposium on System Identification 2003, Rotterdam, 

18: 549-559.

19: \end{center}

20: \section{Introduction}

21: Smoothing Spline ANOVA (SS-ANOVA) models in reproducing kernel

22: Hilbert spaces (RKHS) provide a very general framework for 

23: data analysis, modeling and learning in a variety 

24: of fields. Discrete, noisy scattered, direct and 

25: indirect observations can be accommodated with 

26: multiple inputs and multiple possibly correlated 

27: outputs and a variety 

28: of meaningful structures. 

29: The purpose of this paper is to 

30: give a brief overview of the approach and 

31: describe and contrast a series of applications, 

32: while noting some recent results.

33: