Intelligent Diagnosis of Rolling Bearing Fault Based on Improved Convolutional Neural Network and LightGBM

Table 2

Value of important parameters of LightGBM.

Names of the parameters	Parameter value
n_estimators	500
max_depth	6
num_leaves	34
learning_rate	0.07
bagging_fractin	0.68
bagging_freq	5
feature_fraction	0.6
lambda_l1，lambda_l2	3.4, 2.1
min_data_in_leaf	33
min_split_gain	0.96
min_sum_hessian_in_leaf	0.003