Classification Performance Analysis of Decision Tree-Based Algorithms with Noisy Class Variable

Table 3

F-Measure for the C4.5/CART/RF algorithms on the data sets at varying noise levels.

Data set	0% noise	10% noise	20% noise	30% noise
Adult	91.13/90.26/91.40	90.90/90.27/90.95	90.63/90.10/90.03	89.32/89.19/87.36
Banknote Authentication	98.78/97.46/99.42	97.70/95.47/98.01	95.49/95.10/93.66	91.82/92.58/84.52
Blood Transfusion	86.25/86.90/84.13	85.78/85.49/84.59	86.33/85.77/81.06	85.43/84.17/77.87
Breast Cancer	96.36/95.72/97.64	94.40/95.85/96.86	94.67/95.55/95.17	94.10/93.83/91.83
Car	82.14/86.03/91.35	78.39/82.37/89.56	68.99/78.50/84.50	65.18/74.38/82.04
CMC	48.27/51.61/52.27	46.65/52.02/50.85	47.58/49.17/47.98	44.07/48.88/46.89
Congressional Voting	97.08/95.86/96.98	95.57/96.14/96.35	95.74/95.50/94.79	91.61/92.95/89.57
Dry Bean	92.27/88.06/93.57	91.18/88.14/93.26	88.59/88.01/92.64	82.44/87.16/91.65
Ionosphere	85.85/81.53/90.21	81.73/78.14/89.75	72.68/74.19/87.14	59.35/69.16/79.74
Iris	94.38/91.20/93.48	91.57/94.67/92.86	89.16/91.35/88.40	82.50/87.66/77.86
Optical Digits	89.88/76.54/98.07	87.74/74.84/97.99	83.86/74.12/98.09	75.04/71.90/97.83
Pen-Based Recognition	95.91/82.77/99.18	94.75/81.09/99.19	93.21/79.14/98.89	87.86/78.80/98.06
Qualitative Bankruptcy	96.99/97.32/100	97.89/97.13/98.70	96.40/97.51/92.87	94.01/90.95/82.52
Raisin	85.79/85.14/85.45	85.03/85.20/84.44	84.13/84.60/80.69	82.03/82.08/73.25
Seeds	91.88/91.86/92.79	86.96/90.17/91.49	84.04/90.08/90.02	80.01/87.04/82.81
Sonar	79.35/73.18/87.79	68.30/71.53/80.98	66.34/70.92/82.53	63.19/57.87/74.55
Tic-Tac-Toe Endgame	77.85/88.29/98.50	73.45/85.87/95.48	65.47/71.51/83.87	60.48/63.41/71.61
Vertebral Column	75.45/80.01/80.76	73.42/75.11/81.11	72.15/70.52/77.96	72.86/70.56/73.79
Wine	91.86/86.65/97.41	79.87/88.69/96.97	78.37/82.33/98.26	66.13/78.88/91.33
Wireless Indoor	97.28/96.48/98.36	96.20/96.29/98.34	95.64/95.94/97.73	88.17/95.39/97.01