Decision tree algorithm for tank damage analysis in combat simulation tests
Zhiwu Tang, Qing Xue, Meng Zhao, Wei Yang · 2009
Computer simulations are being seen as the third mode of science, complementing theory and experiment. The vast quantities of data that simulation system produce, gather and store fuel the demand for more effective means of deriving value from them. Data mining technique can discover the valuable knowledge from the massive data automatically. Through analyzing the massive simulation data by data mining technique, the deeply hidden knowledge can be discovered. By means of the combat simulation system of the digitized armored battalion was introduced, the algorithm of decision tree was analyzed. Large amounts of simulation tests were carried out according to the operational solution of the armored forces' landing attack. After generalizing the acquired data, the decision tree was constructed by means of the ID3 algorithm. Finally, 12 rules were extracted from the decision tree, and the further analysis of the results was presented.