The deep learning-based latent factor model discussed in the Deep learning for latent factor collaborative filtering section can be designed as illustrated in Figure 6.4:
Figure 6.4: Deep learning-based latent factor model on Movie Lens 100 K dataset
The user_ID and the movie_ID picks up the user and the movie embedding vector from their corresponding embedding matrices. In the diagram, embedding_1 represents the embedding layer for the user IDs, while embedding_2 represents the embedding layer for movie IDs. The dot product of the user embedding vector and the movie embedding vector is performed ...
