Skip to content

Rank-one updates and other potential performance gains for CUR #216

Open
@agoscinski

Description

@agoscinski

This is a revive of the draft PR #86 (please look into it for further information) because I think it is worth to look into this more given that CUR outperforms FPS by far in regression quality and is often not used because it is so expensive to compute.

The core idea is to update the eigenvectors after a selection instead of recomputing them by an eigendecomposition. @ceriottm mentioned in a discussion that it was mathematically unstable for eigenvectors corresponding to degenerated eigenvalues. So this deserves some dedicated time look into this in detail.

Links:

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or requestlow-prioritySomething is not so important

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions