Abstract:

Decision tree is a usual method of classification in data mining. In this paper, a new heuristic function to build decision trees based on the variable precision rough set is proposed for the inconsistency in the employment of university graduates. The measure of quality of classification acts as an information function to select the condition attribute in this method, and the condition attribute is to be the decision tree node to divide the data set. The dependency and redundancy between attributes are considered; especially a certain inconsistency is allowed to exist in the examples of the positive regions. The method classifies the  data of employment correctly and finds some valuable results for analysis and decision, and it  simplifies the decision trees and improves the extensive ability of decision rules.

Key words: decision tree;variable precision rough set;employment of university graduates;decision rules