Dyah Adila

UW-Madison CS Dept. adila@wisc.edu.

prof_pic.jpg

Hello! I am currently a PhD student at UW-Madison, advised by Fred Sala.

I am broadly interested in how to build efficient machine learning (ML) solutions that are robust and reliable. I am also interested in exploring the rationales behind ML failure modes (e.g., distribution shift, capturing long-term dependency structure, etc.)

During my master’s at the University of Minnesota, I was fortunate to be advised by Ju Sun. I also worked closely with Dongyeop Kang and with Shashi Shekhar.

Originally, I came from Indonesia and finished my undergrad at NTU Singapore. In my free time, I enjoy reading investigative journalism books and swimming in nature.

news

Jun 14, 2022 Our paper Causal Omnivore: Fusing Noisy Estimates of Spurious Correlations is accepted to Workshop on Spurious Correlations, Invariance, and Stability at ICML’22! ! :sparkles: :smile:
May 16, 2022 Our paper Liger: Shoring Up the Foundations: Fusing Model Embeddings and Weak Supervision is accepted as an oral at UAI’22! :tiger:
Sep 3, 2021 Joined UW-Madison as a PhD student :grin:

selected publications

  1. ICML
    Causal Omnivore: Fusing Noisy Estimates of Spurious Correlations
    Adila, Dyah, Cromp, Sonia, Mo, Sicheng, and Sala, Frederic
    In ICML’22 Workshop on Spurious Correlations, Invariance, and Stability Jun 2022
  2. UAI
    Shoring Up the Foundations: Fusing Model Embeddings and Weak Supervision
    Chen, Mayee F, Fu, Daniel Y, Adila, Dyah, Zhang, Michael, Sala, Frederic, Fatahalian, Kayvon, and Ré, Christopher
    In Proceedings of the conference on Uncertainty in artificial intelligence Aug 2022
  3. NeurIPS
    Understanding Out-of-distribution: A Perspective of Data Dynamics
    Adila, Dyah, and Kang, Dongyeop
    In ICBINB@NeurIPS 2021 Nov 2021