Sunday, 15 September 2013

Database supporting fast approximate nearest neighbor queries

Database supporting fast approximate nearest neighbor queries

Is there a database that supports fast approximate nearest neighbor
queries in high-dimensional vector spaces?
I'm looking for a database that would fit the following use case:
Works for millions of points
Works for hundreds-thousands of dimensions
Potentially uses cover trees or locality sensitive hashing for indexing
Does a robust implementation of this exist?

No comments:

Post a Comment