ABSTRACT
In the remainder of the book, tree structured procedures will be developed and studied in a general framework, which includes regression, classification, and class probability estimation as special cases. The reader is presumed to be familiar with the motivation and intuition concerning tree structured procedures. But otherwise the material in the following chapters can be read independently of that given earlier. In Chapters 9 and 10 the joint distribution of X and Y is assumed to be known. Additional topics that arise when tree structured procedures are applied to a learning sample will be dealt with in Chapter 11. Chapter 12 is devoted to a mathematical study of the consistency of tree structured procedures as the size of the learning sample tends to infinity.
