How the RecommenderNet model works?

33 Views Asked by At

I'm implementing a RecommenderNet model with data of 100 users, 27 locations, and 1170 reviews. I want to know how the model works why dot product is used between user_vector and movie_vector, like whether the model uses User-based or Item-based. And I followed an article on kaggle and saw that they used an Embedding Size of 100 compared to the default of 50 and the learning rate changed from 0.001 to 0.004. So for my relatively small data, which parameter is better to use?

0

There are 0 best solutions below