Data Mining in the Medical Domain

This book deals with the Ac4ft-Miner data mining procedure, implemented in the LISp-Miner system, which has been developed at the Department of Information and Knowledge Engineering at the University of Economics, Prague. This procedure is based on association rules; however, it finds the rules that express which actions should be performed to improve the defined state (e.g. to improve the successfulness of a medical treatment). The first goal of the paper is to describe the procedure in a simple and understandable way. The second goal is to apply the procedure to the medical data and present examples of the usage of the procedure. Finally, I aim to create a usage methodology for doctors. Chapter 1 characterises the overall process of knowledge discovery in databases. Chapter 2 presents theoretical concepts related to the Ac4ft-Miner. Chapter 3 deals with action rules. Chapter 4 addresses possibilities for defining the input and interpretation of the output of the Ac4ft-Miner. Chapter 5 describes the research conducted on a real medical data set and gives examples of the tasks solved by the procedure. Finally, chapter 6 offers the Ac4ft-Miner usage methodology for doctors.

Autorentext

Bc. et Bc. Viktor Nekvapil: Studied International Trade and Informatics at the University of Economics, Prague. Currently, he studies Information Technologies at the same university. E-mail: