Application of decision-tree-based algorithm in aeroengine test data mining
Peng Xing-hui · Machinery Design and Manufacture · 2005
How to supplement the missing data correctly is very important in the data pretreatment process This paper describes an new algorithm based on decision tree to solve this problem. The algorithm constructs decision tree using an improved ID3 algorithm, and fills the missing data by decision rules .Before constructing decision tree, attribute reduction method was adopted to obtain the condition attribute collection , and it made the decision tree much simple and the computational speed more fast. The algorithm was applied to analyze an aeroengine test database , and the results showed that it's feasible and high efficient.