Study on Application of Bayesian Classifier Model in Data Stream

Traditional data classification algorithms can not be directly applied for unlimited data and concept drift problem of data stream, so it is accordingly proposed a real-time streaming data classification algorithm for data stream with the concept drift. Bayesian classifier algorithm for the concept drift of stream data summarize the data statistically within the time window then reorganize data set according to the weight of each time window, finally generate a single Bayesian classifier based on the new data set. Experimental results show that the algorithm performance advantages in the classification, classification accuracy and speed.