Using Trajectory Subclustering to Improve Destination Prediction

Algorithm 1

	inputs:, a set of trajectories, , a parameter matrix
	outputs:, the set of clusters extracted from all trajectories in
(1)
	//for each row in M
(2)	fordo
(3)	ifthen
(4)	break
(5)	end
(6)
(7)
(8)	ifthen
(9)
(10)
(11)	else
(12)
(13)	fordo
	//for each initial cluster, generate a set of sub-clusters
(14)
(15)
(16)
	//add each sub-cluster to the final output
(17)
(18)	end
(19)
(20)	end
(21)	end
	//return the set of clusters
(22)	return