Introduction to Statistical Modeling in R

This chapter explores what a statistical model is, R objects which build models, and the basic R notation, called formulas used for models.

In this chapter, you'll start building models: specifying what variables models should relate to one another and training models on the available data. You'll also provide new inputs to models to generate the corresponding outputs.

This chapter is about techniques for deciding whether an explanatory variable improves the prediction performance of a model. You'll use cross validation to compare different models.

This chapter is about constructing models to explore masses of data, for instance to generate hypotheses about what factors are important in how a system works. You'll see how the recursive partitioning model architecture, which has an internal logic for selecting explanatory variables, can be used to explore potentially complex relationships among variables. The chapter also covers the evaluation of prediction performance in models where the response variable is categorical, that is, models used for classification.

Real-world systems are complicated. To faithfully reflect that complexity, models can incorporate multiple explanatory variables. This chapter introduces the notion of covariates and how they allow you to model the effect of an explanatory variable while taking into account the effects of other variables.