Skip to main content

Table 1 Optimized model hyper-parameters

From: AZOrange - High performance open source machine learning for QSAR modeling in a graphical programming environment

Algorithm Parameter Range
RF The number of active attributes
SVM C 2-5 to 215
  γ 2-15 to 23
ANN The number of hidden neurons (one-layer)
CvBoost Maximum branching depth 1 to 20
  The number of trees 1 to 1000
PLS The number of components 1 to nAttr-1
  1. Model hyper-parameters being optimized by default and their corresponding ranges.
  2. 1nAttr is the number of attributes in the data set
  3. 2nEx is the number of examples in the data set