Online Machine Learning from Non-stationary Data Streams in the Presence of Concept Drift and Class Imbalance: A Systematic Review
In IoT environment applications generate continuous non-stationary data streams with in-built problems of concept drift and class imbalance which cause classifier performance degradation.The imbalanced data affects the classifier during concept detection and concept adaptation.In general, for concept detection, a separate mechanism is added in para