Class Selection Avoiding Key-like Property for Better Precise Rules

When we want to select a class attribute among several choices, we suggest to using a method that is based on the overall correctness of a generating rule set as well as the number of different values in the candidate class attributes to avoid key-like property to be selected. The underlying theoretical background for this decision is rough set theory. By scoring the factors among possible candidate class attributes, we may get a better rule set of precision.

• Type: Conference
• Year: 2006
• Where: DMIN
• Authors: Hyontai Sug
• Keywords: Candidate Class, Class Attributes, DMIN 2006, DMIN 2007, Rule Set