Masha Naslidnyk

Hi! I am a PhD student at the Fundamentals of Statistical Machine Learning research group and the Foundational AI CDT at University College London, advised by F-X Briol, Jeremias Knoblauch, and Carlo Ciliberto. Prior to starting my PhD, I was a Machine Learning Scientist at Amazon Research in Cambridge, where I worked on Alexa question answering (2015-2019), and then on Gaussian processes for supply chain emulation (2019-2021). I graduated from Part III in Pure Mathematics at the University of Cambridge in 2014.

My research interests lie broadly in the topics in Gaussian processes and kernel methods; at present, I am focussed on robust inference in conditional probability models.

I will happily respond to “Masha”, but if you’d like to pronounce my last name, it’s nah-sleed-nyk.

news

Jul 01, 2024	💬 Giving a talk at the 2024 ISBA World Meeting in Venice, Italy.
Sep 01, 2023	🌍 Visiting CISPA Helmholtz Center for Information Security under the Helmholtz Visiting Researcher Grant, September-November 2023.
May 08, 2023	💬 Co-organising the Distance-based methods in Machine Learning workshop.
Mar 29, 2023	💬 Our paper Optimally-Weighted Estimators of the Maximum Mean Discrepancy for Likelihood-Free Inference was accepted at ICML 2023.
Jan 12, 2023	🎉 An upcoming paper, Robust Empirical Bayes for Gaussian Processes, won an ASA Section on Bayesian Statistical Science (SBSS) student paper award. It will be presented at JSM 2023.

papers

* indicates equal contribution.

UAI

Conditional Bayesian Quadrature

Zonghao Chen^*, Masha Naslidnyk^*, Arthur Gretton, and François-Xavier Briol

In Uncertainty in Artificial Intelligence, 2024

arXiv Bib PDF Code

@inproceedings{chen2024conditional,
  title = {Conditional Bayesian Quadrature},
  author = {Chen, Zonghao and Naslidnyk, Masha and Gretton, Arthur and Briol, Fran{\c{c}}ois-Xavier},
  booktitle = {Uncertainty in Artificial Intelligence},
  year = {2024},
  organization = {PMLR},
  selected = true,
}

SIAM/ASA JUQ

Comparing Scale Parameter Estimators for Gaussian Process Interpolation with the Brownian Motion Prior: Leave-One-Out Cross Validation and Maximum Likelihood

Masha Naslidnyk, Motonobu Kanagawa, Toni Karvonen, and Maren Mahsereci

To appear in SIAM/ASA Journal on Uncertainty Quantification, 2024

arXiv Bib PDF

@article{naslidnyk2023comparing,
  title = {Comparing Scale Parameter Estimators for Gaussian Process Interpolation with the Brownian Motion Prior: Leave-One-Out Cross Validation and Maximum Likelihood},
  author = {Naslidnyk, Masha and Kanagawa, Motonobu and Karvonen, Toni and Mahsereci, Maren},
  journal = {To appear in SIAM/ASA Journal on Uncertainty Quantification},
  year = {2024},
  selected = true,
}

ICML

Optimally-Weighted Estimators of the Maximum Mean Discrepancy for Likelihood-Free Inference

Ayush Bharti, Masha Naslidnyk, Oscar Key, Samuel Kaski, and François-Xavier Briol

In International Conference on Machine Learning, 2023

arXiv Bib PDF Code

@inproceedings{bharti2023optimally,
  title = {Optimally-Weighted Estimators of the Maximum Mean Discrepancy for Likelihood-Free Inference},
  author = {Bharti, Ayush and Naslidnyk, Masha and Key, Oscar and Kaski, Samuel and Briol, François-Xavier},
  booktitle = {International Conference on Machine Learning},
  year = {2023},
  organization = {PMLR},
  selected = true,
}

NeurIPS

Invariant Priors for Bayesian Quadrature

Masha Naslidnyk, Javier Gonzalez, and Maren Mahsereci

In NeurIPS 2021 Workshop. Your Model Is Wrong: Robustness and Misspecification in Probabilistic Modeling, 2021

arXiv Bib PDF

@inproceedings{naslidnyk2021invariant,
  title = {Invariant Priors for Bayesian Quadrature},
  author = {Naslidnyk, Masha and Gonzalez, Javier and Mahsereci, Maren},
  booktitle = {NeurIPS 2021 Workshop. Your Model Is Wrong: Robustness and Misspecification in Probabilistic Modeling},
  year = {2021},
  selected = true,
}

EMNLP-IJCNLP

Using Pairwise Occurrence Information to Improve Knowledge Graph Completion on Large-Scale Datasets

Esma Balkır, Masha Naslidnyk, Dave Palfrey, and Arpit Mittal

In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), 2019

arXiv Bib PDF Code

@inproceedings{balkir2019using,
  title = {Using Pairwise Occurrence Information to Improve Knowledge Graph Completion on Large-Scale Datasets},
  author = {Balk{\i}r, Esma and Naslidnyk, Masha and Palfrey, Dave and Mittal, Arpit},
  booktitle = {Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)},
  year = {2019},
  selected = true,
}

NeurIPS

Improving knowledge graph embeddings with inferred entity types

Esma Balkır, Masha Naslidnyk, Dave Palfrey, and Arpit Mittal

In NeurIPS 2018 Workshop: Relational Representation Learning, 2018

Bib PDF

@inproceedings{balkir2018improving,
  title = {Improving knowledge graph embeddings with inferred entity types},
  author = {Balk{\i}r, Esma and Naslidnyk, Masha and Palfrey, Dave and Mittal, Arpit},
  booktitle = {NeurIPS 2018 Workshop: Relational Representation Learning},
  year = {2018},
  selected = true,
}