Recurrent Adaptive Classifier Ensemble for Handling Recurring Concept Drifts

Algorithm 1

Framework of RACE.

	Input: () chunks of streaming data
	M: a set of diverse models previously learned
	Output :: the generalized ensemble model at time step t
(1)	For each data chunk do
(2)	Learn a new base model with
(3)	Select transferred models by transferring the highly diverse stored models M
(4)	Build the generalized ensemble using the transferred models and the newly learned model
(5)	Update M with to maximise diversity
(6)	Endfor