Table 8 Performance of different methods versus proposed GCN-ARE for Houston Dataset.

class	GAT²¹	ViT²⁵	Hybrid²³	GEN¹⁹	SCCAN¹⁰	EfficientFormer²⁶	GCN-ARE
1	92.15	89.57	90.68	96.71	90.48	90.10	99.60
2	90.71	98.74	98.91	86.58	99.16	98.66	96.89
3	96.76	100.00	97.47	93.33	100.00	97.33	95.98
4	98.66	99.74	99.39	99.10	99.56	99.74	93.41
5	93.94	97.98	98.06	98.00	98.28	97.97	99.60
6	94.24	67.68	85.67	100.00	100.00	98.99	100.00
7	67.61	88.83	87.99	76.86	88.24	85.49	92.98
8	88.43	97.65	94.28	88.38	90.93	91.06	82.56
9	76.97	87.26	84.67	80.20	79.70	82.38	90.34
10	77.51	95.60	94.23	68.32	95.20	93.55	96.25
11	71.66	84.97	85.92	63.75	79.68	83.66	88.99
12	81.87	93.71	91.90	80.63	90.65	89.80	80.54
13	43.65	83.33	75.86	58.01	78.23	71.76	89.13
14	90.78	91.26	98.75	76.94	96.35	96.57	99.77
14	100.00	98.58	99.52	99.67	99.52	99.21	98.48
OA	82.57	92.53	92.39	83.21	91.52	91.35	92.88
AA	85.19	93.47	93.36	85.25	92.49	92.33	93.63
Kappa	0.81	0.92	0.92	0.82	0.91	0.91	0.92

Quick links

Search