Skip to main content

Advertisement

Table 7 Comparison of methods by maximum Az value using 22 features (Mass)

From: AdaBoost-based multiple SVM-RFE for classification of mammograms in DDSM

  T MGH WU WFUSM SHH
SVM   0.88805 0.93642 0.92474 0.94998
SVM-RFE   0.88849 0.94173 0.93037 0.94998
ENSEMBLE   0.81490 0.90299 0.80317 0.86155
JOIN (1)   0.86728 0.92278 0.87638 0.90789
JOIN (2)   0.83034 0.93886 0.89597 0.85132
JOIN (3)   0.75098 0.87312 0.82694 0.83834
JOIN (4)   0.74270 0.74262 0.66948 0.83834
JOIN (5)   0.68776 0.71316 0.66948 0.80802
MSVM-RFE (bootstrap) 5 0.89720 0.93729 0.92664 0.95087
  10 0.88833 0.93666 0.92972 0.95016
  15 0.89920 0.93746 0.93000 0.95076
  20 0.89014 0.94290 0.92986 0.95066
MSVM-RFE (boost) 5 0.88993 0.93987 0.93581 0.94998
  10 0.88805 0.94315 0.92812 0.94998
  15 0.89092 0.94204 0.92789 0.94998
  20 0.88805 0.94197 0.92758 0.95245
  1. Numbers in parenthesis stands for cutoff value for JOIN method.