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

Delta learning doesn't work in Mean-Approx mode #1

Open
eric-tramel opened this issue Mar 27, 2014 · 0 comments
Open

Delta learning doesn't work in Mean-Approx mode #1

eric-tramel opened this issue Mar 27, 2014 · 0 comments
Assignees
@eric-tramel
Labels
bug

Comments

@eric-tramel
Copy link
Owner

When using a full, explicit, matrix, ample gives correct performance in the following cases:

  • Known noise and known prior parameters.
  • Known noise and unknown prior parameters.
  • Unknown noise and known prior parameters.
  • Unknown noise and unknown prior parameters.

However, when a mean-approximation is used for the F2(...) and F2T(...) operations, delta learning breaks down, working in the following cases:

  • Known noise and known prior parameters.
  • Known noise and unknown prior parameters.
  • Unknown noise and known prior parameters.
  • Unknown noise and unknown prior parameters.

This may relate to the manner in which Delta is being updated.
@eric-tramel eric-tramel self-assigned this Mar 27, 2014
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@eric-tramel eric-tramel

Labels
bug
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@eric-tramel