Short-Term Demand Forecast of E-Commerce Platform Based on ConvLSTM Network

Table 1

Parameter configuration of LightGBM sales prediction model.

Parameter	Setting
boosting_type	GBDT
objective	Regression
metric	MSE
nthread	5
learning_rate	0.01
num_leaves	25
max_depth	15
max_bin	255
subexample	0.8
co1example_bytree	0.8
feature_fraction	0.9
lambda_ll	0.1
lambda_12	0.0