Edwin V. Bonilla

Senior Principal Research Scientist, CSIRO's Data61.

prof_pic.jpg

edwin.bonilla [at] data61.csiro.au

“It’s not [only] the consequence that makes a problem important, it is that you have a reasonable attack.”
Richard Hamming, 1986.

I am a senior principal research scientist and Science Leader for Foundational Machine Learning of the Analytics and Decision Sciences Research program at CSIRO’s Data61, Australia. I also holds an Honorary Associate Professor position at the Australian National University (ANU). Before joining CSIRO, I was a Senior Lecturer at the University of New South Wales (UNSW, 2014-2018), a Senior Researcher at National ICT Australia (NICTA, 2010-2014) and a Research Associate at The University of Edinburgh (UK, 2007-2009).

I obtained a PhD in Informatics at the University of Edinburgh in 2008 and a MSc in Artificial Intelligence with Distinction at the University of Edinburgh (2004). Most of my expertise is in probabilistic modelling and inference algorithms for the analysis of complex data, in areas such as scalable Bayesian inference, Gaussian processes and multi-task learning. I have worked in applications such as geophysical inversions, spatio-temporal modelling, computer vision and document analysis. My current interests include causal discovery and inference, Gaussian processes, Bayesian optimization, optimal design of experiments, neural differential equations and graph neural networks.

My research vision is that of making modern probabilistic machine learning a fundamental part of decision making. This vision is driven by challenging problems in causal discovery and inference and the optimal design of experiments, and underpinned by principled modelling, quantification and propagation of uncertainty. The output of my research is focused on top machine learning venues such as NeurIPS, ICML and AISTATS.

news

latest posts

highlighted publications

  1. Contextual directed acyclic graphs
    Ryan Thompson, Edwin V Bonilla, and Robert Kohn
    In International Conference on Artificial Intelligence and Statistics (AISTATS) , 2024
  2. Variational DAG Estimation via State Augmentation With Stochastic Permutations
    Edwin V Bonilla, Pantelis Elinas, He Zhao, Maurizio Filippone, Vassili Kitsios, and Terry O’Kane
    arXiv preprint arXiv:2402.02644, 2024
  3. Optimal Transport for Structure Learning Under Missing Data
    Vy Vo, He Zhao, Trung Le, Edwin V Bonilla, and Dinh Phung
    Accepted for publication at International Conference on Machine Learning (ICML), 2024
  4. ProDAG: Projection-induced variational inference for directed acyclic graphs
    Ryan Thompson, Edwin V Bonilla, and Robert Kohn
    arXiv preprint arXiv:2405.15167, 2024
  5. Parameter Estimation in DAGs from Incomplete Data via Optimal Transport
    Vy Vo, Trung Le, Long-Tung Vuong, He Zhao, Edwin Bonilla, and Dinh Phung
    In Accepted for publication at International Conference on Machine Learning (ICML) , 2024