Building Recommender System in PHP8
summary
Shows that PHP8 runs real ML pipelines using Rubix ML to load Iris and MovieLens CSVs, vectorize tags and ratings, and run K‑Nearest Neighbors (cosine/euclidean) for collaborative and content‑based recommendations. Demonstrates a concrete dataset (≈600 users, ~10,000 movie features; an example user rated 300) and computes nearest neighbors to produce top‑10 recommendations. Explains dimensionality reduction with PCA/SVD (Netflix Prize 2006–2008, $1M reward, winning solution ~10% better) and builds per‑movie models (≈10,000) to fill sparse matrices. Results are practical and working.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:13Recommendation history
- 3:40Machine learning basics
- 7:26PHP machine learning tools
- 9:01Vectors and datasets
- 17:13Data classification with KNN
- 21:29Content-based vs collaborative
- 29:30Dimensionality reduction
- 35:25Resources and conclusion
related talks
46mBeyond Bag of Words A Practitioner’s Guide to Advanced NLP
01h 21mUnlocking Personalization: A Deep Dive into Modern Recommendation Algorithms
37mA primer on recommendation systems.
01h 31mHands on - Build a Recommender system
33mPicking What to Watch Next: Build a Recommendation System
15mWhat they didn’t teach you in Grad School #bbuzz
48mCollective Intelligence
58mRecommender Systems – A Walkthrough