Hands on - Build a Recommender system
summary
Hands-on build of a movie recommendation engine that actually implements matrix factorization with SVD and a gradient-descent factorizer, uses the Surprise library for grid search and cross‑validation, and runs on ~600 users and 10,000 movies with roughly 100,000 ratings. It shows why filling missing entries with zeros fails, why 3 epochs underfits, how regularization and baselines help cold‑start, and produces a working predictive matrix while calling out scalability and hyperparameter tuning as necessary next steps.
AI-written summary · about
chapters (12)
- 0:00Recommender systems intro
- 2:29What is a recommender
- 4:18Types of recommenders
- 8:24Why collaborative filtering
- 13:22Matrix factorization basics
- 21:02Hands-on setup
- 31:20User-item matrix
- 48:31Gradient descent for MF
- 1:00:08Implementing factorization
- 1:13:10Generating recommendations
- 1:19:06Surprise library
- 1:24:11Summary and next steps
related talks
32mLet's talk about Recommendation Systems!
58mRecommender Systems – A Walkthrough
33mPicking What to Watch Next: Build a Recommendation System
37mA primer on recommendation systems.
01h 38mBuilding Recommender Systems Using Python
01h 21mUnlocking Personalization: A Deep Dive into Modern Recommendation Algorithms
57mHands-On with LLM-Powered Recommenders: Hybrid Architectures
17mLarge-Scale Parallel Recommender System in Julia