Table 6. SA3C-ID model training parameters.

Parameters	Value
Number of pigeons $NU{M}_{pigeon}$	128
Number of iterations $NU{M}_t$	100
Number of annealing iterations $NU{M}_{at}$	100
Weighting factor of Pearson ${\omega }_1$	0.4
Weighting factor of Cosine ${\omega }_2$	0.4
Weighting factor of Jaccard ${\omega }_3$	0.2
Weighting factor of number of selections $\lambda$	0.0075
Weighting factor of attenuation factor $\beta$	0.8
Learning rate of SA3C-ID	0.01
Mini-Batch buffer size $|M|$	200
Discount factor $\gamma$	0.001
Entropy regularization coefficient $\alpha$	0.2
Number of epoch	100
Number of worker	4
Experience replay pool size	1,500