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

Table 4

Classification accuracy comparison for personality prediction of DeepLSTM model with other state-of-the-art algorithms on the ASCERTAIN and the proposed EEG dataset.

Dataset	Method	Validation technique	Accuracy
			Max	Avg	Min
ASCERTAIN	ANN	50–50	70.84	67.24	64.48
	KNN	50–50	67.64	64.26	62.23
	LIBSVM	50–50	77.28	73.84	71.86
	HGP	50–50	78.52	75.38	72.68
	DeepLSTM classifier	50–50	82.48	80.36	77.42
ASCERTAIN	ANN	60–40	74.34	69.86	67.74
	KNN	60–40	70.38	68.16	65.68
	LIBSVM	60–40	79.86	77.28	75.46
	HGP	60–40	81.27	78.73	76.08
	DeepLSTM classifier	60–40	88.14	85.63	81.46
ASCERTAIN	ANN	70–30	75.18	73.16	69.94
	KNN	70–30	72.82	70.84	68.62
	LIBSVM	70–30	83.26	81.62	79.86
	HGP	70–30	86.64	83.38	80.74
	DeepLSTM classifier	70–30	92.86	89.68	86.46
ASCERTAIN	ANN	10-fold	78.82	74.64	72.46
	KNN	10-fold	74.36	72.37	70.25
	LIBSVM	10-fold	84.82	82.42	80.28
	HGP	10-fold	86.12	83.86	81.84
	DeepLSTM classifier	10-fold	95.32	94.16	91.98
Proposed dataset	ANN	50–50	72.84	69.36	66.48
	KNN	50–50	69.32	66.92	63.86
	LIBSVM	50–50	79.74	76.64	73.86
	HGP	50–50	80.36	77.83	74.89
	DeepLSTM classifier	50–50	84.56	82.44	79.62
Proposed dataset	ANN	60–40	76.16	73.28	70.12
	KNN	60–40	72.64	70.62	68.54
	LIBSVM	60–40	81.26	79.58	77.82
	HGP	60–40	83.28	80.94	78.63
	DeepLSTM classifier	60–40	91.52	87.62	84.86
proposed dataset	ANN	70–30	78.62	73.94	71.28
	KNN	70–30	74.68	72.36	70.82
	LIBSVM	70–30	85.64	83.70	81.58
	HGP	70–30	87.69	84.74	82.84
	DeepLSTM classifier	70–30	94.82	91.68	88.72
Proposed dataset	ANN	10-fold	80.24	76.54	74.98
	KNN	10-fold	76.82	74.64	72.28
	LIBSVM	10-fold	88.04	85.16	83.26
	HGP	10-fold	90.32	86.93	84.78
	DeepLSTM classifier	10-fold	96.94	95.88	93.94

Optimal values are represented in bold.