A Cumulants-Based Human Brain Decoding

Table 2

Accuracy comparison of various methods using the fMRI data set.

Condition	LDA (%)	Deep learning (%)	Lib SVM (%)	LRBSF (%)	Proposed method (%)
Human vs animal	58.10	60.60	59.50	61.15	66.98
Human vs building	62.40	64.14	63.27	64.92	70.43
Human vs natural scenes	65.95	67.94	67.60	68.37	71.84
Human vs fruit	62.84	67.20	64.13	67.79	69.96
Animal vs building	65.25	68.80	66.23	69.42	71.71
Animal vs natural scenes	68.84	70.95	67.49	71.30	73.66
Animal vs fruit	66.12	67.17	67.74	67.73	68.88
Building vs natural scenes	58.54	60.10	57.68	59.47	67.72
Building vs fruit	62.29	64.25	63.93	64.62	70.10
Natural scenes vs fruit	62.47	64.84	63.92	65.52	69.82
Average	63.25	65.60	64.10	66.10	70.11