Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[FEA] Handle sparse input for SVM #2197

Open
tfeher opened this issue May 5, 2020 · 2 comments
Open

[FEA] Handle sparse input for SVM #2197

tfeher opened this issue May 5, 2020 · 2 comments
Assignees
@mfoerste4
Labels
CUDA / C++ CUDA issue feature request New feature or request inactive-30d inactive-90d

Comments

@tfeher
Copy link
Contributor

tfeher commented May 5, 2020

Is your feature request related to a problem? Please describe.
cuML's SVM implementation currently uses dense arrays. It works reasonable well even if the number of nonzeros are only 20% of the input data, but it is not ideal for very sparse problems.

Describe the solution you'd like
Allow sparse format for the feature matrix X.

Additional context
The feature matrix enter only during the kernel function calculation K(x_i, x_j), therefore the required changes should be reasonable well confined.
@tfeher tfeher added ? - Needs Triage Need team to review and classify feature request New feature or request labels May 5, 2020
@Garfounkel Garfounkel removed the ? - Needs Triage Need team to review and classify label May 5, 2020
@teju85 teju85 added the CUDA / C++ CUDA issue label May 8, 2020
@github-actions
Copy link

This issue has been labeled inactive-90d due to no recent activity in the past 90 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed.

@github-actions
Copy link

This issue has been labeled inactive-30d due to no recent activity in the past 30 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed. This issue will be labeled inactive-90d if there is no activity in the next 60 days.

@mfoerste4 mfoerste4 mentioned this issue Feb 22, 2023
Merged
@mfoerste4 mfoerste4 mentioned this issue Mar 15, 2023
Merged
rapids-bot bot pushed a commit to rapidsai/raft that referenced this issue Apr 25, 2023
@mfoerste4
Gram matrix support for sparse input (#1296)
1defccc 
This PR adds sparse input support (CSR) for GramMatrix kernel computation. This is a requirement to enable SVM support for sparse input in [cuML issue 2197](rapidsai/cuml#2197). 

It also adds row norm computation for CSR which is utilized for expanded L2 norm computation within RBF kernels.

Although this branch introduces a new API it is still backwards compatible with the old GramMatrix API (which is marked as deprecated).

CC @cjnolet @tfeher

Authors:
  - Malte Förster (https://github.com/mfoerste4)
  - Corey J. Nolet (https://github.com/cjnolet)

Approvers:
  - Tamas Bela Feher (https://github.com/tfeher)
  - Corey J. Nolet (https://github.com/cjnolet)

URL: #1296
ahendriksen pushed a commit to ahendriksen/raft that referenced this issue Apr 27, 2023
@mfoerste4 @ahendriksen
Gram matrix support for sparse input (rapidsai#1296)
909e2c3 
This PR adds sparse input support (CSR) for GramMatrix kernel computation. This is a requirement to enable SVM support for sparse input in [cuML issue 2197](rapidsai/cuml#2197).

It also adds row norm computation for CSR which is utilized for expanded L2 norm computation within RBF kernels.

Although this branch introduces a new API it is still backwards compatible with the old GramMatrix API (which is marked as deprecated).

CC @cjnolet @tfeher

Authors:
  - Malte Förster (https://github.com/mfoerste4)
  - Corey J. Nolet (https://github.com/cjnolet)

Approvers:
  - Tamas Bela Feher (https://github.com/tfeher)
  - Corey J. Nolet (https://github.com/cjnolet)

URL: rapidsai#1296
rapids-bot bot pushed a commit that referenced this issue Jun 1, 2023
@mfoerste4
Support sparse input for SVC and SVR (#5273)
20fcb7e 
This PR adds support for sparse input to SVR and SVC. 'fit' as well as 'predict' can be called with sparse data compatible/convertible to SparseCumlArray. Support vectors in the model might also be stored as sparse data and can be retrieved as such.
This PR requires rapidsai/raft#1296 to provide sparse kernel computations.
Corresponding issue: #2197

Authors:
  - Malte Förster (https://github.com/mfoerste4)
  - Tamas Bela Feher (https://github.com/tfeher)

Approvers:
  - Tamas Bela Feher (https://github.com/tfeher)
  - Corey J. Nolet (https://github.com/cjnolet)

URL: #5273
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@mfoerste4 mfoerste4

Labels
CUDA / C++ CUDA issue feature request New feature or request inactive-30d inactive-90d
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
4 participants
@teju85 @tfeher @Garfounkel @mfoerste4