Skip to main content

Table 2 Model hyperparameter configurations of machine learning models

From: Prediction of small-molecule compound solubility in organic solvents by machine learning algorithms

Machine learning algorithms Hyperparameter configurations
lightGBM 0.1503; 486; 0.5338; 0.2431; 32; 18
SVM Radial basis function; 112.4; 0.0006046
RF 486; 203; 45; 4; 2
ET 396, 504, 39, 4, 2
kNN 4; the standard Euclidean distance
DT 25; 8; 3
PLS 22
DNN 10; 1024, 256, 128, 64, 64, 64, 64, 64, 64, 64; 0.001; Adam with β1, β2 of 0.9, 0.999; 100; 500; 0.0003
Ridge regression 1
  1. “;” Separates different hyperparameters
  2. “,” The hyperparameter is composed of more than one element