From Data to Recommendations: Building an Intelligent System with ROR
summary
Claims a reproducible pipeline that turns user data into recommendations using Ruby on Rails, shows a working proof: linear regression house-price demo trained for 500 epochs and collaborative-filtering via ALS matrix factorization that fills sparse user–item matrices and returns ranked items. Covers content-based book examples, reports a ~23% lift cited from Nvidia, lists concrete mitigations for cold start, sparsity and scalability, and recommends privacy tools like federated learning and differential privacy.
AI-written summary · about
chapters (14)
- 0:00Building recommendation systems
- 1:04AI ML and deep learning basics
- 2:05ML demo with Ruby on Rails
- 4:31Generative modeling and neural networks
- 7:43ML pipeline stages
- 8:45What are recommendation systems
- 10:30Logic behind algorithms
- 12:35Collaborative filtering explained
- 13:59Matrix factorization with ALS
- 15:45Content-based filtering demo
- 17:16Key challenges in systems
- 19:24Mitigation strategies
- 20:06Importance and future of systems
- 21:31Resources and conclusion
related talks
55mBuilding a Recommendation Engine with Machine Learning Techniques
33mPeople who liked this talk also liked ... Building Recommendation...
58mRecommender Systems – A Walkthrough
38mRecommendations at Rappi: from MVP to Data Product
35mIntelligent Ruby: Getting Started with Machine Learning
37mA primer on recommendation systems.
32mLet's talk about Recommendation Systems!
29mThe Challenges of Building a Personalized Recommendation System in E-Commerce