A More Formal Treatment of Classification and Forecasting

Abstract

This chapter covers much of the foundational material from the last chapter but more formally and in more detail. The chapter opens with a discussion of the model used to characterize how the data were generated. That model is very different from the one used in conventional regression. Then, classification is considered followed by the estimation issues it raises. The bias-variance tradeoff is front and center. Some material in this chapter may require careful reading.