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

Table 3

Learning rules variations for the Mackey-Glass and Duffing time series.

Learning rule	Mackey-Glass time series			Duffing time series
	(𝐾=10 step ahead prediction)			(𝐾=10 step ahead prediction)
	MSE	NMSE	𝑟	MSE	NMSE	𝑟
Conjugate gradient	0.00545	0.09321	0.94965	0.12332	0.80365	0.77838
Momentum	0.004475	0.07940	0.9596	0.00335	0.02185	0.98910
Quick propagation	0.00534	0.09486	0.9481	0.00660	0.04305	0.97821