Algorithms of Nonmonotonic Data Mining Based on Concept Hierarchy and Layered Mining

杨杰,叶晨洲,李国正
DOI: https://doi.org/10.3321/j.issn:1006-2467.2001.11.012
2001-01-01
Abstract:Few existing data mining techniques consider nonmonotony in data mining. In this paper, nonmonotony in data mining was dealt with by concept hierarchy and layered mining. An algorithm for extracting nonmonotonic rules based on rough set and concept hierarchy and an algorithm for nonmonotonic data mining based on hybridized knowledge representation models and layered mining were presented. By an example of fault diagnosis the principle of nonmonotonic data mining based on hybridized knowledge representation models and layered mining was described. And an algorithm for automatic generation of fuzzy rules, and the construction of a diagnostic neural network were presented.
What problem does this paper attempt to address?