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.

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

Atom domain FrankWolfe Algorithm for vector problems #1087

Merged
merged 27 commits into from Sep 5, 2017

Conversation

Projects
None yet
4 participants
@czdiao
Contributor

czdiao commented Aug 9, 2017

No description provided.

@czdiao

This comment has been minimized.

Show comment
Hide comment
@czdiao

czdiao Aug 9, 2017

Contributor

Hi! This is the FrankWolfe Algorithm for vector atom domain problems. I added the new functionalities of:

  1. Regularization
  2. Support Prune
  3. Line search method for classic FrankWolfe
  4. New Atoms class interface
  5. Structured group constraint type
  6. Full Corrective Update rule. (Update with atom norm constraint.)
Contributor

czdiao commented Aug 9, 2017

Hi! This is the FrankWolfe Algorithm for vector atom domain problems. I added the new functionalities of:

  1. Regularization
  2. Support Prune
  3. Line search method for classic FrankWolfe
  4. New Atoms class interface
  5. Structured group constraint type
  6. Full Corrective Update rule. (Update with atom norm constraint.)
@czdiao

This comment has been minimized.

Show comment
Hide comment
@czdiao

czdiao Aug 11, 2017

Contributor

Hi! The windows builds here failed. I didn't see any error messages. Could anybody help me to check it? Thanks! @stephentu @rcurtin @zoq

Contributor

czdiao commented Aug 11, 2017

Hi! The windows builds here failed. I didn't see any error messages. Could anybody help me to check it? Thanks! @stephentu @rcurtin @zoq

czdiao added some commits Aug 28, 2017

@rcurtin

Looks nice, I pointed out a few comments. I am curious, do we have any machine learning methods where these optimizers could be applied, or an example of how these optimizers can be used in practice? It might be helpful to add that information for a user. (I don't know if maybe you have a tutorial you are planning to add.)

@stephentu

This comment has been minimized.

Show comment
Hide comment
@stephentu

stephentu Aug 30, 2017

Contributor

@rcurtin: an application we are working on is matrix completion, to augment/replace the current finicky SDP solver. that will be a separate PR though.

Contributor

stephentu commented Aug 30, 2017

@rcurtin: an application we are working on is matrix completion, to augment/replace the current finicky SDP solver. that will be a separate PR though.

czdiao added some commits Aug 31, 2017

@stephentu stephentu merged commit 17aa9ce into mlpack:master Sep 5, 2017

4 checks passed

Static Code Analysis Checks Build finished.
Details
Style Checks Build finished.
Details
continuous-integration/appveyor/pr AppVeyor build succeeded
Details
continuous-integration/travis-ci/pr The Travis CI build passed
Details
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment