About

I’m a PhD student at Columbia, advised by professor John Cunningham. My research interests include statistical machine learning, data mining, and bioinformatics. Prior to Columbia, I studied Mathematics and Statistics at Oxford.

Preprints and Publications

See a more up to date list on Google Scholar.

  • Elliott Gordon-Rodriguez, Thomas Quinn, and John Cunningham. “Data Augmentation for Compositional Data: Advancing Predictive Models of the Microbiome.” Neural Information Processing Systems (NeurIPS), 2022 (to appear). [paper]

  • Elliott Gordon-Rodriguez*, Gabriel Loaiza-Ganem*, Andres Potapczynski, and John Cunningham. “On the Normalizing Constant of the Continuous Categorical Distribution.” arXiv, 2022. [paper]

  • Elliott Gordon-Rodriguez, Gabriel Loaiza-Ganem, and John Cunningham. “On a novel probability distribution for zero-laden compositional data.” Proceedings of CodaWork, 2022 (to appear). [paper]

  • Elliott Gordon-Rodriguez, Thomas Quinn, and John Cunningham. “Fast log-ratio selection for high-dimensional compositional data.” CodaWork Abstracts, 2022 (to appear). [link]

  • Elliott Gordon-Rodriguez. “Advances in Machine Learning for Compositional Data.” Doctoral Dissertation, Columbia University, 2022. [paper]

  • Elliott Gordon-Rodriguez. “On Disentanglement and Mutual Information in Semi-Supervised Variational Auto-Encoders.” Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2021. [paper]

  • Thomas Quinn, Elliott Gordon-Rodriguez, and Ionas Erb. “A critique of differential abundance analysis, and advocacy for an alternative.” arXiv, 2021 (Submitted). [paper]

  • Elliott Gordon-Rodriguez, Thomas Quinn, and John Cunningham. “Learning Sparse Log-Ratios for High-Throughput Sequencing Data.” Bioinformatics, 2021. [paper]

  • Elliott Gordon-Rodriguez, Gabriel Loaiza-Ganem, Geoff Pleiss and John Cunningham. “Uses and Abuses of the Cross-Entropy Loss: Case Studies in Modern Deep Learning.” In 1st I Can’t Believe It’s Not Better Workshop, Neural Information Processing Systems (NeurIPS), 2020. [paper] [talk]

  • Elliott Gordon-Rodriguez, Gabriel Loaiza-Ganem, and John Cunningham. ‘‘The Continuous Categorical: a Novel Simplex-Valued Exponential Family." International Conference on Machine Learning (ICML), 2020. [paper] [talk]

Software

I have built and maintain R and Python implementations of CoDaCoRe, which can be found on CRAN and Github, respectively. For any questions or comments related to these, feel free to drop me a line at eg2912@columbia.edu, or see the package documentation and Guide.