Class Central is learner-supported. When you buy through links on our site, we may earn an affiliate commission.

YouTube

Graph-Based Approximate Nearest Neighbors and HNSW

Pinecone via YouTube

Start learning Write review
Coursera Ad

AI, Data Science & Cloud Certificates from Google, IBM & Meta

Learn More → University of the People Ad

Earn Your Business Degree, Tuition-Free, 100% Online!

Learn More →

Overview

Google, IBM & Meta Certificates – 40% Off
One Coursera Plus subscription covers most Professional Certificates on Coursera.
Unlock All Certificates
This practical workshop explains graph-based approximate nearest neighbor search, focusing on HNSW and related graph indexes. It covers construction, memory-constrained scenarios, updates, deletions, coarse quantization, disk-based search, and benchmarking.

Syllabus

Intro
Vector Search
Exhaustive Search
Approximate Search
Many ANNS Algorithms
Graph algorithms
Advantages of graph algorithm
Delaunay graphs and Voronoi diagrams
Problems with Delaunay graphs
Delaunay Graph Subgraphs
Relative neighborhood graph (RNG)
Skip-lists analogy
HNSW construction
Extension to memory-constrained scenarios
Using graphs a coarse quantizer (ivf-hnsw)
DiskANN
SPANN and HNSW-IF
Updates and deletions.
Benchmarking SQUAD
Benchmarking MSMARCO
Practical advice

Taught by

Pinecone

Reviews

Start your review of Graph-Based Approximate Nearest Neighbors and HNSW

Start learning

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Sign up for free
Someone learning on their laptop while sitting on the floor.