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

Table 1

Optimal Parameters of FTLRNN for both time series.

Sr. no.	Parameters	Hidden layer for MG	Hidden layer for Duffing	Output layer for MG	Output layer for Duffing
1	Number of PEs	38	21	1	1
2	Transfer function	Tanh	Tanh	Tanh	Lin tanh
3	Learning rule	Momentum	Momentum	Momentum	Momentum
4	Step size	1	1	0.1	0.1
5	Momentum	0.8	0.8	0.8	0.8