Many data sources produce new data points at a very high rate. With so much data, the issue of data quality emerges. Low quality data can degrade the accuracy of machine learning models that are built around those data sources. Ideally, we would have completely clean data sources, but that’s not very realistic. One alternative