Data Mining Algorithms: Explained Using R

Chapter 1Tasks

1.1 Introduction

This chapter discusses the assumptions and requirements of the three major data mining tasks this book focuses on: classification, regression, and clustering. It adopts a machine learning perspective, according to which they are all instantiations of inductive learning, which consists in generalizing patterns discovered in the data to create useful knowledge. This perfectly matches the predictive modeling view of data mining adopted by this book, according to which the ultimate goal of data mining is delivering models applicable to new data. While the book also discusses tasks that are not directly related to model creation, their only purpose is to make the latter easier, more reliable, and more effective. These auxiliary tasks—attribute transformation, discretization, and attribute selection—are not discussed here. Their definitions are presented in the corresponding chapters.

Inductive learning is definitely the most commonly studied learning scenario in the field of machine learning. It assumes that the learner is provided with training information (usually—but not necessarily—in the form of examples) from which it has to derive knowledge via inductive inference. The latter is based on discovering patterns in training information and generalizing them appropriately. The learner is not informed and has no possibility to verify with certainty which of the possible generalizations are correct and can only be hoped, but never guaranteed to succeed. ...