Flow Graphs as a Tool for Mining Prediction Rules of Changes of Components in Temporal Information Systems

摘要

Flow Graphs proposed by Z. Pawlak are a very useful tool for reasoning from data. In this paper, an application of flow graphs is presented. We use them to mine rules describing component changes in the consecutive time windows which a given temporal information system is split into. Such rules can be helpful for predicting future changes of components in the system and what follows, predicting the future behavior of the analyzed system.