Richard Nock

Research Scientist, Google Research (Brain team) & Honorary Professor (Level E), the Australian National University

email: [ firstlast@google.com (work) | first.m.last@gmail.com (not work) ]


Background, Research interests and more

I defended a PhD in computer science under the supervision of Olivier Gascuel and an accreditation to lead research (HDR) in computer science with Michel Habib. I graduated simultaneously in computer science and in agronomical engineering ("Ingénieur Agro", majors in statistics and industrial microbiology), after two years of CPGE ("Classes Préparatoires aux Grandes Ecoles"). My interests cover Machine Learning, Privacy, Information Geometry at large.

I am an External Advisor to the Australian Artificial Intelligence Institute, and Area Chair for NeurIPS 2021. I was Area Chair for ICML 2021, 2020, 2019, ICLR 2021, AAAI 2021, AISTATS 2019 and NeurIPS 2020, 2019, and a Local Chair of ICML 2017.


Some invited talks material

  • Confidential Computing - Federate Private Data Analysis
    NIPS 2016 PMPML workshop on Private Multi-Party Machine Learning
    [ slides | PMPMLw' page ]


Some recent publications [ Scholar | DBLP ]

  • Zac Cranko, Zhan Shi, Xinhua Zhang, Richard Nock and Simon Kornblith
    Generalized Lipschitz Regularisation Equals Distributional Robustness
    ICML 2021, International Conference on Machine Learning
    [ paper | supplementary | poster, code, slides, video: soon ]
  • Richard Nock, Stephen Hardy, Wilko Henecka, Hamish Ivey-Law, Jakub Nabaglo, Giorgio Patrini, Guillaume Smith and Brian Thorne
    The Impact of Record Linkage on Learning from Feature Partitioned Data
    ICML 2021, International Conference on Machine Learning
    [ paper | supplementary | poster, code, slides, video: soon | FLOW seminar ]
  • Amir Dezfouli, Richard Nock and Peter Dayan
    Adversarial Vulnerabilities of Human Decision-Making
    Proceedings of the National Academy of Sciences USA, 2020 (Vol 117), pp 2922129228
    [ paper (open access) | supplementary | data | code ]
  • Richard Nock and Frank Nielsen
    The Phylogenetic Tree of Boosting has a Bushy Carriage but a Single Trunk
    Proceedings of the National Academy of Sciences USA, 2020 (Vol 117), pp 8692-8693
    [ paper (open access) ]
  • Christian Walder and Richard Nock
    All your Loss are Belong to Bayes
    NeurIPS 2020, Advances in Neural Information Processing Systems
    [ paper | supplementary | poster | code | video ]
  • Richard Nock and Aditya Krishna Menon
    Supervised Learning: No Loss No Cry
    ICML 2020, International Conference on Machine Learning
    [ paper | supplementary | video ]
  • Christian Simon, Piotr Koniusz, Richard Nock and Mehrtash Harandi
    Adaptive Subspaces for Few-Shot Learning
    CVPR 2020, IEEE International Conference on Computer Vision and Pattern Recognition
    [ paper | code ]
  • Christian Simon, Piotr Koniusz, Richard Nock and Mehrtash Harandi
    On Learning to Modulate the Gradient for Fast Adaptation of Neural Networks
    ECCV 2020, European Conference on Computer Vision
    [ paper | code ]
  • Hisham Husain, Borja Balle, Zac Cranko and Richard Nock
    Local Differential Privacy for Sampling
    AISTATS 2020, International Conference on Artificial Intelligence and Statistics
    [ paper | supplementary | code ]
  • Richard Nock, Natalia Polouliakh, Frank Nielsen, Keiko Oga, Carlin R. Connell, Cedric Heimhofer, Kazuhiro Shibanai, Samik Ghosh, Ken-ichi Aisaki, Satoshi Kitajima, Jun Kanno, Taketo Akama and Hiroaki Kitano
    A Geometric Clustering Tool (AGCT) to Robustly Unravel the Inner Structure of Time-Series Gene Expressions
    PLoS ONE, 2020 15(7): e0233755
    [ paper | AGCT access ]
  • Hisham Husain, Richard Nock and Robert C. Williamson
    A Primal-Dual link between GANs and Autoencoders
    NeurIPS 2019, Advances in Neural Information Processing Systems
    [ paper | supplementary | poster ]
  • Amir Dezfouli, Hassan Ashtiani, Omar Ghattas, Richard Nock, Peter Dayan and Cheng Soon Ong
    Disentangled Behavioral Representations
    NeurIPS 2019, Advances in Neural Information Processing Systems
    [ paper | supplementary | poster | BioRxiv version ]
  • Soumava Kumar Roy, Mehrtash Harandi, Richard Hartley and Richard Nock
    Siamese Networks: the Tale of Two Manifolds
    ICCV 2019, International Conference on Computer Vision (Oral)
    [ paper | supplementary | video | code ]
  • Richard Nock and Robert C. Williamson
    Lossless or Quantized Boosting with Integer Arithmetic
    ICML 2019, International Conference on Machine Learning
    [ paper | supplementary | poster | code | slides ]
  • Zac Cranko and Richard Nock
    Boosted Density Estimation Remastered
    ICML 2019, International Conference on Machine Learning
    [ paper | supplementary | poster | code | slides ]
  • Zac Cranko, Aditya Krishna Menon, Richard Nock, Cheng Soon Ong, Zhan Shi and Christian Walder
    Monge blunts Bayes: Hardness Results for Adversarial Training
    ICML 2019, International Conference on Machine Learning
    [ paper | supplementary | ArXiv | poster | code | slides ]
  • Samitha Herath, Basura Fernando, Mehrtash Harandi and Richard Nock
    Min-Max Statistical Alignment for Transfer Learning
    CVPR 2019, IEEE International Conference on Computer Vision and Pattern Recognition
    [ paper | supplementary | poster | code ]
  • Amir Dezfouli, Bernard Balleine and Richard Nock
    Optimal Response Vigor and Choice under Non-Stationary Outcome Values
    Psychonomic Bulletin and Review 2019 (Vol 26), pp 182-204
    [ paper ]
  • Marta Avalos-Fernandez, Richard Nock, Cheng Soon Ong, Julien Rouar and Ke Sun
    Representation Learning of Compositional Data
    NeurIPS/NIPS 2018, Advances in Neural Information Processing Systems
    [ paper + supplementary | poster | code | video ]
  • Amir Dezfouli, Edwin Bonilla and Richard Nock
    Variational Network Inference: Strong and Stable with Concrete Support
    ICML 2018, International Conference on Machine Learning
    [ paper | supplementary | poster | code | slides ]
  • Kelvin Hsu, Richard Nock and Fabio Ramos
    Hyperparameter Learning for Conditional Kernel Mean Embeddings with Rademacher Complexity Bounds
    ECML-PKDD 2018, European Conference on Machine Learning and KDD (Best Student ML Paper Award)
    [ paper + supplementary | poster | code | slides | video ]
  • Richard Nock, Zac Cranko, Aditya Krishna Menon, Lizhen Qu and Robert C. Williamson
    f-GANs in an Information Geometric Nutshell
    NIPS 2017, Advances in Neural Information Processing Systems (Spotlight)
    [ paper | supplementary | ArXiv (longer) | poster | code | slides | video ]
  • Giorgio Patrini, Alessandro Rozza, Aditya Krishna Menon, Richard Nock and Lizhen Qu
    Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approach
    CVPR 2017, IEEE International Conference on Computer Vision and Pattern Recognition (Oral)
    [ paper | code | slides | video ]
  • Boris Muzellec, Richard Nock, Giorgio Patrini and Frank Nielsen
    Tsallis Regularized Optimal Transport and Ecological Inference
    AAAI 2017, AAAI Conference on Artificial Intelligence
    [ paper | supplementary | poster | code ]
  • Richard Nock
    On Regularizing Rademacher Observation Losses
    NIPS 2016, Advances in Neural Information Processing Systems
    [ paper | supplementary | poster | code | video ]
  • Richard Nock, Aditya Krishna Menon and Cheng Soon Ong
    A scaled Bregman Theorem with Applications
    NIPS 2016, Advances in Neural Information Processing Systems
    [ paper | supplementary | poster | video]
  • Richard Nock, Raphael Canyasse, Roksana Boreli and Frank Nielsen
    k-variates++: more Pluses in the k-means++
    ICML 2016, International Conference on Machine Learning
    [ paper | supplementary | poster | slides ]
  • Giorgio Patrini, Frank Nielsen, Richard Nock and Marcello Carioni
    Loss Factorization, Weakly Supervised Learning and Label Noise Robustness
    ICML 2016, International Conference on Machine Learning
    [ paper | supplementary | poster | slides ]
  • Giorgio Patrini, Richard Nock, Stephen Hardy and Tiberio Caetano
    Fast Learning from Distributed Datasets without Entity Matching
    IJCAI 2016, International Joint Conference on Artificial Intelligence
    [ paper | ArXiv (longer) | poster | slides ]
  • Frank Nielsen, Boris Muzellec and Richard Nock
    Classification with Mixtures of Curved Mahalanobis Metrics
    ICIP 2016, International Conference on Image Processing (Oral)
    [ paper | supplementary | slides ]
  • Richard Nock, Frank Nielsen and Shun-ichi Amari
    On Conformal Divergences and their Population Minimizers
    IEEE Trans. on Information Theory 2016 (Vol 62), pp 527-538
    [ paper | ArXiv (paper + supplementary + additional results) ]

Students

    PhD program (completion year):

  • 2021: Hisham Husain (ANU)
  • 2020: Zac Cranko (ANU)
  • 2016: Giorgio Patrini (ANU)

    Interns:

  • 2019: Ahmed Kriouile (Ecole Polytechnique)
  • 2019: Yannis Bekri (Ecole Polytechnique)
  • 2019: Paul Roussel (Ecole Polytechnique)
  • 2018: Mona Buisson-Fenet (Ecole des Mines de Paris)
  • 2018: Olivia Torres (Ecole des Mines de Paris)
  • 2016: Boris Muzellec (Ecole Polytechnique)
  • 2015: Raphael Canyasse (Ecole Polytechnique)
  • 2015: Alexis Le Dantec (Ecole Polytechnique)