Forecasting Foreign Exchange Volatility Using Deep Learning Autoencoder-LSTM Techniques

Table 5

Five error measures for each model in the JYVIX.

	MSE	RMSE	MAE	MPE (%)	MAPE (%)
Period 1
LSTM	0.6152	0.7843	0.4611	−0.4230	4.6438
Autoencoder-LSTM	0.4359	0.6602	0.4430	−1.3870	4.5153
Period 2
LSTM	0.3321	0.5763	0.4525	−0.2644	3.5579
Autoencoder-LSTM	0.4261	0.6528	0.4933	0.79419	3.9160
Period 3
LSTM	3.8273	1.9563	1.0422	−5.3252	15.0497
Autoencoder-LSTM	2.7176	1.6485	0.9447	2.6840	13.4985