Announcing Trustworthy ML Initiative and Virtual Seminar Series (starts Oct 29)

8 views
Skip to first unread message

shf...@gmail.com

unread,
Oct 22, 2020, 7:17:51 AM10/22/20
to Crowdsourcing and Human Computation
(Apologies for cross posting)

Hi all,

We wanted to share an initiative we launched recently: Trustworthy ML Initiative (TrustML). A major focus of the initiative is a bi-weekly virtual seminar series where speakers discuss their work in various subfields of Trustworthy ML, such as explainability, fairness, differential privacy, causality, robustness, etc.

The first seminar starts next Thursday Oct 29, 9am PT / 5pm London / 7pm Addis Ababa. We would like to invite you to attend the seminar and participate in the post-seminar discussion!

Our confirmed speakers include Percy Liang, Ayanna Howard, Irene Chen, Jenn Wortman Vaughan, Cynthia Rudin, Pin-Yu Chen, Zachary Lipton, Steven Wu, Shibani Santurkar, Celia Cintas, Katherine Heller, Gautam Kamath, Suresh Venkatasubramanian, Sherri Rose, Alexander D'Amour, and more to come!

As part of our seminar series, we are also featuring students in Rising Star Spotlight Talks. Please contact us at trustw...@gmail.com if you are a student and want to present your work.

To enable easy access of fundamental resources in the field, we have also collected links to courses, textbooks, videos, etc. on our website. We also manage an active Twitter account @trustworthy_ml that disseminates the latest work in trustworthy ML.

We welcome you to engage with our resources, attend our seminars, and send us your work to be disseminated. All feedback is welcome!

On behalf of the Trustworthy ML Initiative organizers and advisors,
Hima Lakkaraju (Harvard)
Sara Hooker (Google Brain)
Sarah Tan (Facebook)
Subho Majumdar (AT&T Labs Research)
Chhavi Yadav (UCSD)
Jaydeep Borkar (Pune University)
Kamalika Chaudhuri (UCSD)
Tom Dietterich (Oregon State University)
Kush Varshney (IBM Research)


************************

More about the Trustworthy ML initiative: As machine learning (ML) systems are increasingly being deployed in real-world applications, it is critical to ensure that these systems are behaving responsibly and are trustworthy. To this end, there has been growing interest from researchers and practitioners to develop and deploy ML models and algorithms that are not only accurate, but also explainable, fair, privacy-preserving, causal, and robust. This broad area of research is commonly referred to as trustworthy ML.

While it is incredibly exciting that researchers from diverse domains ranging from machine learning to health policy and law are working on trustworthy ML, this has also resulted in the emergence of critical challenges such as information overload and lack of visibility for work of early career researchers. Furthermore, the barriers to entry into this field are growing day-by-day -- researchers entering the field are faced with an overwhelming amount of prior work without a clear roadmap of where to start and how to navigate the field.

To address these challenges, we are launching the Trustworthy ML Initiative (TrustML) with the following goals:

- Enable easy access of fundamental resources to newcomers in the field.
- Provide a platform for early career researchers to showcase and disseminate their work.
- Encourage discussion and debate on the latest work on trustworthy ML.
- Develop a community of researchers and practitioners working on topics related to trustworthy ML.

Please check our website www.trustworthyml.org for our ongoing efforts and programs!
Reply all
Reply to author
Forward
0 new messages