In recent years, there have been some interesting studies on predictive modeling in data streams. However, most such studies assume relatively balanced and stable data streams but cannot handle well skewed (e.g., few positives but lots of negatives) and skewed distributions, which are typical in many data stream applications. In this paper, we propose an ensemble and cluster based sample method to deal with this situation. The study shows that this method has effective result on skewed data streams mining.
Accurate Quantile Estimation for Skewed Data Streams Using Nonlinear Interpolation