Improved method for data reinforcement based on ROUSTIDA
Hongxia Wang · Journal of Naval University of Engineering · 2011
For the deficiency in ROUSTIDA,an improved method was proposed.Firstly,in dealing with the condition attributes and decision attributes in different ways,priority was given to the decision attributes to avoid generating the inconsistent decision system.Then the missing condition attri-bute values were filled under the prerequisite of the same decision attribute value to expand the application of the original algorithm.Finally,an example was used to show that the improved method can get more centralized decision rules.