EEG-Based Personality Prediction Using Fast Fourier Transform and DeepLSTM Model

Table 5

Comparison of sensitivity, precision, and specificity of DeepLSTM model for various partition schemes.

Dataset	Validation technique	Sensitivity (%)		Precision (%)		Specificity (%)		F1 score (%)
		Mean ± Std		Mean ± Std		Mean ± Std		Mean ± Std
ASCERTAIN	50–50	81.64	±3.08	80.53	±3.12	78.42	±2.84	80.56	±2.36
	60–40	87.76	±3.14	86.82	±3.24	85.94	±2.68	86.44	±2.44
	70–30	91.49	±3.16	90.56	±3.42	89.67	±2.36	90.40	±3.42
	10-fold	94.72	± 3.16	93.48	± 3.12	92.86	± 2.98	93.68	± 2.84
Proposed dataset	50–50	83.46	±3.14	82.75	±3.15	81.14	±3.24	82.46	±3.18
	60–40	90.74	±3.24	89.25	±3.16	88.94	±3.18	89.46	±3.32
	70–30	93.54	±3.14	92.85	±3.25	91.94	±3.42	92.28	±3.24
	10-fold	95.86	± 3.18	94.44	± 3.14	93.84	± 3.12	94.96	± 3.16

Optimal values are represented in bold.