GENERALIZED ADDITIVE MODELS

INTRODUCTION

Introduction

EXAMPLE

Illustration of GAMs

THE METHOD

How do GAMs work?

Regression models

	Classical linear
	Generalized linear
	Additive
	Generalized additive

Categorical predictors

Smoothing methods

	Running means
	Running medians
	Running lines
	Kernel
	Splines
	Loess smoother

Multiple predictors

Degrees of smoothing

	Constructing a neighbourhood
	Span sizes

ROLE OF GAMs

Uses of GAMs

Advantages / Disadvantages

EXAMPLE

Application to spinal surgery data

Stages in fitting a generalized additive model

	Preliminary examination
	Choosing a link function
	Examining predictors
	Selecting predictors

Stages in fitting a parametric model

	Start + age
	Age
	Start
	GAM v parametric

CONCLUSION

Conclusion

APPLICATION SOFTWARE

Application software

	S-Plus
	Genstat

References

Glossary