Syllabus
- Introduction to Recommender Systems - mission, requirements, methods, data
- User Feedback
- Collaborative filtering, KNN, matrix factorization methods
- Content-based filtering
- representation of ordering, Fagin-Lotem-Naor (FLN) model
- graphical form of FLN, Challenge-response framework
- Hybrid and context-aware recommender systems
- Evaluation of recommender systems, real-world applications
Annotation
Due to the extreme information overload on the web, we need models that can process information in a personalized way. One class of such models are Recommender Systems (RS).
The core of RS are machine learning algorithms focusing on user feedback. RS aim to predict users’ future preferences and provide them with surprising, yet relevant objects.
This course covers common working principles of recommender systems, its learning methods, data types, requirements and evaluation as well as some aspects of the practical deployment.