continuous machine learning pipelines are common in industrial settings where
models are periodically trained on data streams. Unfortunately, concept drifts
may occur in data streams where the joint distribution of the data X and label
y, P(X, y), changes over time and possibly degrade
这篇论文介绍了一种新的集成学习方法,称为 Diversity and Transfer based Ensemble Learning(DTEL),用于处理概念漂移的增量学习。通过利用保留的历史模型和转移学习,DTEL 可以更有效地处理概念漂移,并通过对 15 个合成数据流和 4 个真实数据流的经验研究证明了其有效性。