Data streams mining demand not only the processing speed under the limited computing resources, but the adaptive processing capacity of concept drift. This paper proposes a ensemble classifier method to process data streams classification, which divides the stream data into blocks to train, and the data sets for each class label are trained within a base classifier. According to an optimization strategy, the data blocks are updated in real time to deal with the concept drift. In the integrated model, the base classifier is updated at the same time by updating the weights of each base classifier. The experimental results verify the superiority of this method.
展开▼
