Accuracy of Classifier Combining Based on Majority Voting
Hong Peng, Chengde Lin, Luo Linkai, Qifeng Zhou · 2007
In this paper, we formulate the accuracy of classifier combining which is based on majority voting, there are only two parameter involved, one is the average accuracy of individual classifiers, the other we call it lapsed accuracy (LA) is related with the efficiency of classifier combining, and we discuss the theoretical bounds of majority voting via the formula.