An ISVM Algorithm Based on High-Dimensional Distance and Forgetting Characteristics

Table 14

The train_ACC values comparison of all the algorithms for different datasets.

Datasets	Train_ACC (%)
	Simple-ISVM	KKT-ISVM	CSV-ISVM	GGKKT-ISVM	CD-ISVM	HDFC-ISVM	HDFC-ISVM
Breast_cancer	94.78	95.0	97.69	96.50	97.85	96.12	98.86
German	94.63	93.23	98.85	95.60	97.70	97.81	99.50
Heart	99.75	98.63	99.60	98.89	98.90	99.57	100
Image	98.42	96.10	99.79	97.56	98.30	98.72	99.47
Mushroom	99.5	98.75	98.86	98.76	98.72	99.71	99.96
Thyroid	97.52	97.78	97.65	97.80	99.40	98.79	99.44
Titanic	76.94	81.98	77.03	81.85	81.83	80.75	82.33
Splice	90.84	92.26	90.19	92.20	92.19	92.26	92.26
Diabetes	94.33	94.73	94.95	94.10	94.20	94.98	97.81
Credit	86.52	85.12	85.60	85.60	85.80	86.01	86.09
Spambase	78.81	79.73	79.05	79.80	79.81	79.72	80.56
Waveform	68.34	65.67	68.97	67.20	68.53	68.97	71.28
Mean	90.03	89.92	90.68	90.49	91.10	91.11	92.29