Dual Policy Learning for Aggregation Optimization in Graph Neural Network-based Recommender Systems

What is it about?

DPAO is a Reinforcement Learning based model that allows us to adaptively choose the number of GNN (Graph Neural Network) layers for users and items, respectively, on GNN-based recommendation tasks. Our model considers the heterogeneous nature of users and items to select the number of layers and achieves promising results on recommendation tasks by attaching our model to the existing GNN-based models.

Featured Image

Photo by Milad Fakurian on Unsplash

Photo by Milad Fakurian on Unsplash