DriftSurf: A Risk-competitive Learning Algorithm under Concept Drift
This work addresses the challenge of maintaining model accuracy under concept drift for applications in streaming data analysis, representing an incremental improvement over existing drift-detection methods.
The authors tackled the problem of concept drift in streaming data by developing DriftSurf, an adaptive learning algorithm that integrates drift detection into a stable-state/reactive-state process to improve detection rates and reduce false positives. Their theoretical analysis shows the algorithm's risk is competitive with an oracle-aware algorithm, and experiments on synthetic and real datasets confirm these findings.
When learning from streaming data, a change in the data distribution, also known as concept drift, can render a previously-learned model inaccurate and require training a new model. We present an adaptive learning algorithm that extends previous drift-detection-based methods by incorporating drift detection into a broader stable-state/reactive-state process. The advantage of our approach is that we can use aggressive drift detection in the stable state to achieve a high detection rate, but mitigate the false positive rate of standalone drift detection via a reactive state that reacts quickly to true drifts while eliminating most false positives. The algorithm is generic in its base learner and can be applied across a variety of supervised learning problems. Our theoretical analysis shows that the risk of the algorithm is competitive to an algorithm with oracle knowledge of when (abrupt) drifts occur. Experiments on synthetic and real datasets with concept drifts confirm our theoretical analysis.