An Improved Sequential Recommendation Algorithm based on Short-Sequence Enhancement and Temporal Self-Attention Mechanism

Table 2

The parameters of the proposed model and the experimental environment.

Model	Learning rate	0.001
	Momentum	0.9
	Dropout rate	0.2
	Batch size	128
	Maximum iterations	200
	Validation interval	20
	Regularization	0.00005
	Short-sequence threshold	20
	Maximum sequence length for ML-1M	70
	Maximum sequence length for AM-BE	30
	Latent dimension for ML-1M	50
	Latent dimension for AM-BE	20
	Pseudo-historical item for ML-1M	5
	Pseudo-historical item for AM-BE	15
Environment	Programming software	Python3.6
	Deep learning framework	Pytorch
	Computer system	Windows 10
	Cpu	E5-2620 v4
	RAM	32.0 GB
	Gpu	GeForce RTX 2080