YujiaBao/tofu

"Learning Stable Classifiers by Transferring Unstable Features" ICML 2022

36
/ 100
Emerging

This project helps machine learning researchers or practitioners develop more reliable classification models by leveraging characteristics from one related domain to improve predictions in another. It takes existing labeled datasets from a 'source' domain and applies its 'unstable features' to enhance a classifier's stability and accuracy on a 'target' domain, resulting in more robust models even when the target data is challenging. This is for researchers and practitioners working on real-world classification tasks.

No commits in the last 6 months.

Use this if you need to build a stable classifier for a specific domain, but you have limited or challenging data and believe a related domain's data could offer useful insights, even if those insights are subtle or 'unstable' in their original context.

Not ideal if you are looking for a general-purpose, off-the-shelf classifier and do not have access to or expertise in leveraging related domain datasets for transfer learning.

machine-learning-research domain-adaptation classification-modeling transfer-learning model-robustness
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 5 / 25
Maturity 16 / 25
Community 15 / 25

How are scores calculated?

Stars

14

Forks

4

Language

Python

License

MIT

Last pushed

Jul 24, 2022

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/YujiaBao/tofu"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.