Abstract
A random forest algorithm with an improved decision tree and an optimal tree integration is proposed. First, an improved ID3 decision tree is proposed. Next, when the trees are integrated, each decision tree is assigned a voting right consistent with the out-of-bag accuracy of the decision tree. Finally, the experimental results show that the improved random forest algorithm is efficient and the classification performance is improved compared with the traditional random forest algorithm.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have