Research Article
Stream Classification Algorithm Based on Decision Tree
Algorithm 1
Classification algorithm DVFDTc.
| (1) | Win1 = S0/∗ Training set to build initial classification model ∗/ | | (2) | M1 = BuildVFDT(Win1) | | (3) | ClassValue = M1.Classify(Win1) | | (4) | Winc = S0, ClassValue; | | (5) | Mc = BuildVFDT(Winc) | | (6) | For di = 1, …, n do | | (7) | ClassValueoi = M1.Classify(di) | | (8) | ClassValueci = Mc.Classify(di) | | (9) | If(ClassValueoi = = ClassValueci) | | (10) | WriteToLocal(ClassValueoi) | | (11) | Wintmp.add(ClassValueoi) | | (12) | Else | | (13) | For j = 1, …, k do | | (14) | Hj = Mix_Sample(Win1,Winc) | | (15) | Mj = buildVFDT(Hj) | | (16) | For j = 1, …, k do | | (17) | Rhj = Mj.classify(di) | | (18) | tmpR = Most_Appears(Rh) | | (19) | WriteToLocal(tmpR) | | (20) | If (i% Winc.size = = 0) | | (21) | Win1 = Winc | | (22) | Win2 = Wintmp | | (23) | M1 = Mc | | (24) | M2 = BuildVFDT(Winc) |
|
