streaming-ml

This paper empirically studies retraining policies for streaming machine learning systems under concept drift, budget, and latency constraints.