Decision Tree Construction Algorithm for Incomplete Information System
Jiajun Chen, Yuanyuan Huang
Abstract
Jiajun Chen, Yuanyuan Huang
Abstract
The decision tree algorithm uses forecasting technology and my reject data containing default values when processing incomplete information. As a result, the accuracy of decision rules is degraded. In this study, a decision tree construction algorithm is created based on rough set model of integrated tolerance relation. This algorithm is subject to the attribute significance as heuristic function to select test attributes, and brings in the concept of prior probability and inhibiting factor of incomplete information system to divide subsets of objects with test attribute value of "*". It effectively avoids the loss of some significant information during the process of creating decision tree, thus promoting accuracy of decision rules. Moreover, the algorithm is characterized by noise immunity.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The decision tree algorithm uses forecasting technology and my reject data containing default values when processing incomplete information. As a result, the accuracy of decision rules is degraded. In this study, a decision tree construction algorithm is created based on rough set model of integrated tolerance relation. This algorithm is subject to the attribute significance as heuristic function to select test attributes, and brings in the concept of prior probability and inhibiting factor of incomplete information system to divide subsets of objects with test attribute value of "*". It effectively avoids the loss of some significant information during the process of creating decision tree, thus promoting accuracy of decision rules. Moreover, the algorithm is characterized by noise immunity.
Key concepts: Decision tree, Computer science, Incremental decision tree, Decision tree learning, ID3 algorithm, Data mining, Rough set, Heuristic