Recommander systems

Objectives

Recommendation systems select the options to be presented to users from a large number of possibilities. For example, they recommend the next video to watch, the next song to listen to, the photos to show… The course will present the models used to build such systems.

– Objective: identify the recommendation model to be applied in a given context
– Objective: implement standard recommendation models
– Objective: measure the quality of a model’s recommendations

Course outline

– Purpose of recommender systems

– State of the art of recommender systems (nearest neighbors, collaborative filtering)

– Evaluation of recommender systems

– Problems encountered by recommender systems and related solutions (cold start, exploration-exploitation trade-offs, etc.)

Prerequisites

R, Python, linear algebra, function optimization, statistical learning