You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
\chapter{RankFromSets: Scalable Set Recommendation with Optimal Recall}
\label{ch:rfs}
\lettrine[image=true,lines=3]{design/I}{n} the previous chapter, we built scalable and performant probabilistic models of likely configurations of interacting atoms in statistical physics systems. Modeling choices were guided by the structure of the underlying patterns of interaction between random variables. As another case study, we turn to recommender systems in this chapter, where the core problem is to model which items a user is likely to interact with. By building the structure of individual datapoints into a probabilistic model of user interaction, and considering the goals of the recommendation task, we develop a scalable, accurate framework for recommending items with attributes.