Beyond IGMC

We extend the state-of-the-art Inductive Graph Matrix Completion recommender system by introducing Graph Normalization and Layer Aggregation variants, and explore the models' potent transfer learning capabilities.