TABLE 3.

Experiment 1 - best determined RNN parameters.

Method	Learning rate	Dropout	Epochs	Number of layers	Layer 1	Layer 2
RNN-HASCA	0.010000	0.050000	60	1	5	6
RNN-SCA	0.008531	0.050000	60	1	11	10
RNN-GA	0.002725	0.190960	60	1	15	5
RNN-PSO	0.004895	0.066932	60	1	15	15
RNN-FA	0.010000	0.200000	60	1	15	8
RNN-BSO	0.010000	0.200000	43	1	15	10
RNN-RSA	0.008858	0.064926	60	1	14	10
RNN-COLSHADE	0.005873	0.144412	60	1	12	8