Interactive Decision Tree Construction for Interval and Taxonomical Data

摘要: Visual data-mining strategy lies in tightly coupling the visualizations and analytical processes into one tool that takes advantage of assets from multiple sources. This paper presents two graphical interactive decision tree construction algorithms able to deal either with (usual) continuous data or interval taxonomical data. They are extensions existing algorithms: CIAD [17] PBC [3]. Both can be used an cooperative mode (with automatic algorithm find best split current node). We have modified corresponding help mechanisms allow them interval-valued attributes. Some results obtained on sets presented methods we create these sets.