jyhong836 Goto Github PK

followers: 72.0 following: 143.0 repos: 43.0 gists: 5.0

Name: Junyuan Hong

Type: User

Company: University of Texas at Austin

Bio: Researcher on Federated Learning, Privacy, Trustworthy ML.

Twitter: hjy836

Blog: https://jyhong.gitlab.io

Hi there, I'm Junyuan Hong 👋

I am a researcher on Machine Learning. My research centers around Privacy-Centric Trustworthy Machine Learning. My vision is to enhance trustworthiness (regarding fairness, robustness and security) under the privacy constraint, e.g., federated learning and differentially-private learning. My research highlights:

[Privacy] in centralized learning with theoretic understanding, analysis tool and empirical algorithms and its financial application in distributed learning.
[Trustworthiness] ⨉ Privacy: Federated Learning and Private Edge-Cloud Collaboration.
- Fair pre-training and federated learning📂.
- Robustness by efficient acquisition against adversarial samples📂 and OoD inference📂.
- Secure learning against backdoor attacks in private-data-free distillation.
- Inclusive learning by efficient on-device model adaptation📂, and private outsourcing training and customizable federated learning📂 for low-end devices.

See my homepage and CV for more information.

Junyuan Hong's Projects

academic-kickstart

Easily create a beautiful website using Academic and Hugo

aif360

A comprehensive set of fairness metrics for datasets and machine learning models, explanations for these metrics, and algorithms to mitigate bias in datasets and models.