Skip to main content
. Author manuscript; available in PMC: 2013 May 1.
Published in final edited form as: Artif Intell Med. 2012 Mar 3;55(1):61–70. doi: 10.1016/j.artmed.2012.01.002

Algorithm 2.

Bootstrapping(D)

Require: D = {(x1, y1), … (xN, yN)}
  1: for i = 1 to N do
  2:  for b = 1 to NFOLDS do
  3:    Si ← Resample(D)
  4:    errorb = 0
  5:    totalb = 0
  6:    if (xi, yi) ∉ Si then
  7:      Mi,b ← BuildModel(Si)
  8:      errori,b ← |Mi,b(xi) − yi|
  9:      errorberrorb + errori,b
10:      totalbtotalb + 1
11:      errorberrorb/totalb
12:    end if
13:  end for
14: end for
15: accuracyb11Ni=1Nerrorb
16: M ← BuildModel(D)
17: accuracyt11Ni=1N|M(xi)yi|
18: return accuracy ← 0.632 × accuracyb + 0.368 × accuracyt