Algorithm Specific Settings

–init_method

parameter initializers

default: initialize parameters according to the original paper

normal: initialize parameters with normal distribution

uniform: initialize parameters with uniform distribution

xavier_normal: initialize parameters with xavier_normal distribution

xavier_uniform: initialize parameters with xavier_uniform distribution

–optimizer

optimization method for training the algorithms

default (optimizer in the original paper)

sgd

adam

adagrad

–early_stop

whether to activate the early-stop mechanism

true

false

–tune_testset

whether to directly tune on testset, and the default value is false

true

false

–factors

the dimension of latent factors (embeddings)

–reg_1

the coefficient of L1 regularization

–reg_2

the coefficient of L2 regularization

–dropout

dropout rate

–lr

learning rate

–epochs

training epochs

–batch_size

batch size for training

–num_layers

number of layers for MLP

–alpha

constant to multiply the penalty terms for SLIM

–elastic

the ElasticNet mixing parameter for SLIM in the range of (0,1)

–pop_n

the preliminary selected top-n popular candidate items to reduce the time complexity for MostPop

–maxk

the number of neighbors to take into account for ItemKNN

–node_dropout

node dropout ratio for NGCF

–mess_dropout

message dropout ratio for NGCF

–kl_reg

the coefficient of KL regularization for Multi-VAE