A Multi-level Weighted Concept Drift Detection Method
Meng Han, zhiqiang Chen, Hongxin Wu, Muhang Li, xilong Zhang · Research Square · 2022
Abstract The concept drift detection method is an online learner. Its main task is to determine the position of drifts in the data stream, so as to reset the classifier after detecting the drift to improve the learning performance, which is very important in practical applications such as user interest prediction or financial transaction fraud detection. A new level transition threshold parameter is proposed, and a piecewise weighting mechanism including "Stable Level-Warning Level-Drift Level" is innovatively introduced in concept drift detection. The instances in the window are weighted in levels, and it is applied to the double sliding window. Based on this, a multi-level weighted drift detection method(MWDDM) is proposed. Especially, two variants which are MWDDM_H and MWDDM_M are proposed basd on Hoeffding inequality and Mcdiarmid inequality respectively. Experiments on artificial datasets show that MWDDM can detect abrupt and gradual concept drift faster than any other comparison algorithms, while maintaining a low false positive ratio and false alarm rate. Experiments on real-world datasets show that MWDDM has the highest classification accuracy in most cases.