Product Quantization for Similarity Search

How to compress and fit a humongous set of vectors in memory for similarity search with asymmetric distance computation (ADC)

Peggy Chang
Towards Data Science
8 min readMay 9, 2022

--

Article cover for “Product Quantization for Similarity Search — How to compress and fit a humongous set of vectors in memory for similarity search with asymmetric distance computation (ADC)”. Author: Peggy Chang
Photo by Markus Winkler on Unsplash

Similarity search and nearest neighbor search are very popular and widely used in many fields. They are used in recommendation systems, in online stores and marketplaces that enable…

--

--