Hands-On Deep Learning with R

Performing exploratory data analysis

Recommender systems seek to use available information and preference pattern data to generate predictions about a particular user's interests.

As a starting point, we can use EDA to identify important patterns and trends in the underlying data to inform our understanding and subsequent analysis:

1. Let's identify the top 10 items based on implicit ratings constructed from user-item interaction data using the following code:

# user-item interaction exploratory data analysis (EDA)item_interactions <- aggregate(    rating ~ title, data = steamdata, FUN = 'sum')item_interactions <- item_interactions[    order(item_interactions$rating, decreasing = TRUE),]item_top10 <- head(item_interactions, 10)kable(item_top10) ...