The course on recommender systems conducted in National Research University - Higher School of Economics (Moscow, Russia). Academic year 2024/2025.
- Wiki page of this course
- Table with grades
- The code materials for each seminars can be found in the corresponding folders
/seminar*. - To download any folder please use this link.
- Recordings of lectures and seminars (coming soon).
- All questions can be asked in the Telegram chat (the invitation link is available only to NRU HSE students)
The final grade is calculated as follows:
0.3 * Home Assignment + 0.3 * Quizzes + 0.4 * Exam
where Home Assignments - 5 home assignments in Jupyter Notebook (max 10 points). Quizzes - 19 weekly quizzes on lecture's and seminars' topics in Google Forms (max 10 points). Exam - oral examination on all topics (max 10 points).
- Introduction to recommender systems
- Similarity (neighborhood) based and linear approaches
- Matrix factorization
- Collaborative filtering
- Content-based models
- Hybrid approaches
- Sequential models for next-item recommendations
- Context-based recommendations
- Models for the next-basket recommendations task
- Autoencoders and variational autoencoders for recommendations
- Graph and knowledge-graph based models
- Interpretability and explainability
- Uplift recommendations
- Multi-task & cross-domain recommendations
- RL in RecSys
- Domain recommendations (multiomodal data)
- A/B testing and multi-armed bandites. Model monitoring
- Large scale RecSys
- Vanilla API service for recommender system
- Additional applied aspects and trends in recommender systems
All content created for this course is licensed under the MIT License. The materials are published in the public domain.