Depth and Width Changeable Network-Based Deep Kernel Learning-Based Hyperspectral Sensor Data Analysis

Table 6

Performance evaluation on different kernel learnings.

Datasets	Indian pines dataset		Pavia University dataset
Methods	OA (%)	KC (%)	OA (%)	KC (%)
RBF-SKL [27]	73.23	63.73	75.43	68.76
POLY-SKL [27]	75.77	66.28	78.08	72.07
Mahal-RBF-SKL [31]	76.92	67.79	76.26	69.87
Mahal-Poly-SKL [31]	77.64	68.62	79.07	73.24
SK-CV-RBF-SKL [30]	67.03	64.23	75.71	69.14
SK-POLY-SKL [30]	69.37	66.96	77.62	71.37
EasyMKL [22]	68.13	65.53	76.82	72.14
SimpleMKL [23]	69.22	66.78	73.56	66.76
SM1MKL [24]	77.34	69.62	79.98	74.34
L2MKL [25, 26]	77.37	74.85	79.67	75.87
NMF-MKL [32]	67.48	64.81	71.57	64.42
KNMF-MKL [32]	68.22	65.63	72.80	65.81
MKL1 [28]	74.23	69.63	78.69	72.11
MKL2 [29]	76.07	72.84	79.24	72.92
QMKL	79.28	75.85	80.96	75.34
Proposed deep	81.53	78.34	83.14	78.76