The Backfitting Algorithm
1. Initialize:
b = average(y)
set f1, f2, .., fp by do regression with y and x1, y and x2, ..., y and xp
2. Iteration:
f1 = S1( y -b - f2 - f3 -..-fp | x1) : S1 denotes a smooth of the residuals-1 against the predictor x1
f2 = S2( y -b - f1 - f3 -..-fp | x2) : S2 denotes a smooth of the residuals-2 against the predictor x2
...
fp = Sp( y -b - f1 - f2 -..-fp-1 | xp) : Sp denotes a smooth of the residuals-p against the predictor xp
3. Repeat 2. until the function don't change
b = average(y)
set f1, f2, .., fp by do regression with y and x1, y and x2, ..., y and xp
2. Iteration:
f1 = S1( y -b - f2 - f3 -..-fp | x1) : S1 denotes a smooth of the residuals-1 against the predictor x1
f2 = S2( y -b - f1 - f3 -..-fp | x2) : S2 denotes a smooth of the residuals-2 against the predictor x2
...
fp = Sp( y -b - f1 - f2 -..-fp-1 | xp) : Sp denotes a smooth of the residuals-p against the predictor xp
3. Repeat 2. until the function don't change
留言
張貼留言