pechatov / recsys_course

The course on recommender systems conducted in National Research University - Higher School of Economics (Moscow, Russia). Academic year 2021-2022.

RecSys course

The course on recommender systems conducted in National Research University - Higher School of Economics (Moscow, Russia). Academic year 2021-2022.

Course on recommender systems.

Week1

• Examples of RecSys models in production.
• Formalization of the ranking (recommender systems) task (2 popular types of tasks, 2 types of data sets).
• Ranking functions (BPR, WARP, RankNET, LambdaRank, LambdaMART).
• Metrics for the quality estimation (Hitrate, Precision @ k, Recall @ k, AP@k, MAP@l, DCG@k, NDCG@k).
• Taxonomy of RecSys approaches ([MF, FM, CF & other general], Content-based [including knowledge graph based, GB for ranking], Context-based, Sequential based, RL, Hybrid [including two-level cascade, blending]).
• Pros / Cons of current approaches.
• Review of the main future directions.
• Problems of the current articles.
• Recommended sources on RecSys.

Week2

• Basic baselines
• Item-based and user-based similarity, similarity metrics.
• Matrix Factorization (SVD et al.)
• Collaborative Filtering (ALS)

Week3

• Content-based & Hybrid recommenders taxonomy
• LightFM model and library
• Example of cascade recommender model (using gradient boosting on the second level)
• Important preprocessing steps
• Cross-validation types for recommender systems