In the neo4j documentation, I saw that the node classification algorithm uses Logistic Regression as a classifier. How does it take node embeddings as a feature? Does it take it as a single feature or n (no. dimensions of embedding ) number of features?
Hello mujtaba.mirza,
If you look at the syntax section (https://neo4j.com/docs/graph-data-science/current/algorithms/ml-models/node-classification/#algorithms-ml-nodeclassification-syntax), you can see the featureProperties
parameter. This is used to specify a list of node properties (multiple features).
One of these features could be an embedding.
Note you only specify the featureProperties
when training your model via gds.alpha.ml.nodeClassification.train
. On prediction, we assume there are exactly the same properties on the nodes as specified during train.