Desktop Survival Guide
by Graham Williams
Todo: UNDER CONSTRUCTION
Todo: Needs Introduction. This chapter introduces the concept of descriptive and predictive (covering classification and regression) data mining, and the associated algorithms, and the associated documentation. Then the following chapters cover each of the algorithms. Do we move the framework from the Introduction to here - I think so.
In this chapter we introduce the concepts of descriptive and predictive Analytics
The following chapters are then devoted to particular algorithms.
Modelling is what people often think of when they think of data mining. Modelling is the process of taking some data (usually) and building a model that reflects that data. Usually the aim is to address a specific problem through modelling the world in some way and from the model develop a better understanding of the world.
There is a bewildering array of tools and techniques at the disposal of the data miner. We can get a better understanding of what is available through categorising the algorithms according to the types of analysis performed. In this chapter we introduce and summarise the broader categories of data mining analysis. Part I then presents, in a systematic manner, many algorithms that are used in data mining and available either freely or else implemented in commercial toolkits.
Much of the terminology used in data mining has grown out of that used in both machine learning and statistics. We identify, for example, two very broad categories of analysis as unsupervised and supervised (as in supervised and unsupervised learning).
We introduce such an ordering to the world of data mining techniques in this chapter. In summary: