selected_features.py

# selected_features.py

# Ordered top 30 features
#SELECTED_FEATURES = ['V17', 'V12', 'V14', 'V10', 'V16', 'V11', 'V9', 'V7', 'V18', 'V4',
#       'V21', 'V3', 'V26', 'V20', 'V2', 'V1', 'V19', 'V6', 'Amount', 'V8',
#       'V15', 'Time', 'V5', 'V22', 'V27', 'V24', 'V28', 'V13', 'V25', 'V23']

# Top 20 features from ordered list above^
SELECTED_FEATURES = ['V17', 'V12', 'V14', 'V10', 'V16', 'V11', 'V9', 'V7', 'V18', 'V4',
       'V21', 'V3', 'V26', 'V20', 'V2', 'V1', 'V19', 'V6', 'Amount', 'V8']

#Top 1 features: AUC = 0.8507 (+/- 0.0473)
#Top 2 features: AUC = 0.9042 (+/- 0.0391)
#Top 3 features: AUC = 0.9225 (+/- 0.0244)
#Top 4 features: AUC = 0.9223 (+/- 0.0249)
#Top 5 features: AUC = 0.9262 (+/- 0.0210)
#Top 6 features: AUC = 0.9280 (+/- 0.0218)
#Top 7 features: AUC = 0.9259 (+/- 0.0239)
#Top 8 features: AUC = 0.9319 (+/- 0.0207)
#Top 9 features: AUC = 0.9310 (+/- 0.0224)
#Top 10 features: AUC = 0.9392 (+/- 0.0182)
#Top 11 features: AUC = 0.9442 (+/- 0.0190)
#Top 12 features: AUC = 0.9451 (+/- 0.0202)
#Top 13 features: AUC = 0.9409 (+/- 0.0215)
#Top 14 features: AUC = 0.9410 (+/- 0.0171)
#Top 15 features: AUC = 0.9410 (+/- 0.0139)
#Top 16 features: AUC = 0.9416 (+/- 0.0169)
#Top 17 features: AUC = 0.9437 (+/- 0.0196)
#Top 18 features: AUC = 0.9479 (+/- 0.0230)
#Top 19 features: AUC = 0.9418 (+/- 0.0224)
#Top 20 features: AUC = 0.9488 (+/- 0.0215) # BEST AUC HERE
#Top 21 features: AUC = 0.9478 (+/- 0.0238)
#Top 22 features: AUC = 0.8804 (+/- 0.1091)
#Top 23 features: AUC = 0.8920 (+/- 0.0731)
#Top 24 features: AUC = 0.8985 (+/- 0.0554)
#Top 25 features: AUC = 0.8386 (+/- 0.1882)
#Top 26 features: AUC = 0.8678 (+/- 0.1347)
#Top 27 features: AUC = 0.9112 (+/- 0.0513)
#Top 28 features: AUC = 0.8763 (+/- 0.1004)
#Top 29 features: AUC = 0.8975 (+/- 0.0788)
#Top 30 features: AUC = 0.8864 (+/- 0.1074)