For similarity search in high-dimensional vector spaces (or ‘HDVSs’) in database system, the performance of most methods generally degrades as dimensionality increases. However, there’s a simple vector approximation scheme, called VA-file, exhibit linear complexity at high dimensionality. In this post, I show the comparison of R-tree and VA-File in HDVSs.

Github Repo

Click here to see the details.

Good luck!