Pre-processing
In most machine learning scenarios, several pre-processing steps are applied to produce data that is amenable to machine learning algorithms. This is also true for graph data. The goal of pre-processing is to provide good features for the learning algorithm. As part of our pipelines we offer adding such pre-procesing steps as node property steps (see Node Classification or Link Prediction).
Some options include:
-
Scale Properties to normalize node features
-
One Hot Encoding to encode string properties as numeric arrays