A Historical Context for Data Streams
This work provides historical insight for researchers in data stream mining, but it is incremental as it reviews existing assumptions without introducing new methods or results.
The paper reviews the historical context of data streams research, explaining how common assumptions in machine learning for streaming data, such as single-pass inspection and real-time prediction, originated from computational resource constraints.
Machine learning from data streams is an active and growing research area. Research on learning from streaming data typically makes strict assumptions linked to computational resource constraints, including requirements for stream mining algorithms to inspect each instance not more than once and be ready to give a prediction at any time. Here we review the historical context of data streams research placing the common assumptions used in machine learning over data streams in their historical context.