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.

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

domain adaptation #79

Open
jovo opened this issue Sep 6, 2020 · 4 comments · May be fixed by #464
Open

domain adaptation #79

jovo opened this issue Sep 6, 2020 · 4 comments · May be fixed by #464
Assignees
Labels
feature a feature we want and know how to do

Comments

@jovo
Copy link
Member

jovo commented Sep 6, 2020

given 2 tasks, map posteriors to each other via LDDMM.

(@jdey4 this is different from what you've been working on, though related)

@jovo
Copy link
Member Author

jovo commented Sep 6, 2020

scikit-image/scikit-image#4390 is the code for doing LDDMM.

@jovo jovo added the feature a feature we want and know how to do label Sep 10, 2020
@jovo
Copy link
Member Author

jovo commented Sep 11, 2020

@tliu68 you want this, right?

@tliu68
Copy link
Member

tliu68 commented Sep 11, 2020

@jovo yes, please

@tliu68
Copy link
Member

tliu68 commented Sep 23, 2020

experiments to run:

  • 2-class classification: R-XOR exp where task1 is mapped onto task2 w/ lddmm

maybe also:

  • run (multitask) vary_task_sample exp on original CIFAR10 data w/ lddmm and compare results to those w/ cross-task

  • map original & rotated CIFAR10 data w/ lddmm and compare results to non-adversarial case

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
feature a feature we want and know how to do
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants