International Journal of Computer Applications
In today's world, data is rapidly and continuously growing and is not constant in nature. There is a problem to deal with such kind of evolving data, as it is impractical to store and process this streaming data. Also, in real world application, the data which is coming is typically noisy, has some missing values, redundant features and thus very large time is wasted to preprocess that data. The time complexity can reduce by selecting only useful features to build model for classification. The proposed system addresses the issue of adaptive preprocessing for streaming data.