Enhancing Stock Price Trend Prediction via a Time-Sensitive Data Augmentation Method

Table 4

Comparison experiments by applying interpolation with diverse balance coefficients to the high-frequency subseries and the original time series.

	Interpolation		Interpolation
	High frequency (ours)		Original time series
Balance coefficient ()	MSE	MAE	MSE	MAE
0.2	4.921	0.903	5.218	0.951
0.4	5.050	0.915	5.458	0.976
0.6	4.917	0.893	5.461	0.977
0.8	5.192	0.932	5.5148	0.9647
No data augmentation			5.106	0.923