# From-Scratch Item-Item k-NN with NumPy: Production-Grade on MovieLens-1M

> Source: https://sukruyusufkaya.com/en/learn/oneri-sistemleri/sifirdan-item-item-knn-numpy-movielens-1m
> Updated: 2026-05-13T13:29:35.218Z
> Category: Öneri Sistemleri
> Module: Module 5: Memory-Based Collaborative Filtering (k-NN)
**TLDR:** Module 5's backbone: production-grade item-item k-NN from scratch on MovieLens-1M. Adjusted cosine + shrinkage, sparse matrix optimizations, offline batch precomputation pattern, top-K neighbor caching, second row in our benchmark.