Novel FTLRNN with Gamma Memory for Short-Term and Long-Term Predictions of Chaotic Time Series

Table 12

For K = 5, training and testing samples variation for FTLRNN on testing data set.

Time series		Mackey-Glass time series			Duffing time series
Training exemplars	Testing exemplars	MSE	NMSE	𝑟	MSE	NMSE	𝑟
10%	75%	0.000501	0.01779	0.99122	0.00465	0.08523	0.95975
20%	65%	0.001006	0.019508	0.9902	0.00162	0.01236	0.99388
30%	55%	0.001196	0.02256	0.98875	0.00150	0.00944	0.99552
40%	45%	0.00144	0.02653	0.98675	0.00139	0.00885	0.99578
50%	35%	0.00167	0.0322	0.9838	0.00170	0.01154	0.99435
60%	25%	0.00227	0.04145	0.97911	0.00202	0.01315	0.99356
70%	15%	0.00371	0.06227	0.9684	0.00149	0.00999	0.99524
80%	05%	0.01089	0.1502	0.9226	0.00435	0.02941	0.98926