English

Making Multi-Axis Models Robust to Multiplicative Noise: How, and Why?

Methodology 2026-03-30 v1 Machine Learning

Abstract

In this paper we develop a graph-learning algorithm, MED-MAGMA, to fit multi-axis (Kronecker-sum-structured) models corrupted by multiplicative noise. This type of noise is natural in many application domains, such as that of single-cell RNA sequencing, in which it naturally captures technical biases of RNA sequencing platforms. Our work is evaluated against prior work on each and every public dataset in the Single Cell Expression Atlas under a certain size, demonstrating that our methodology learns networks with better local and global structure. MED-MAGMA is made available as a Python package (MED-MAGMA).

Keywords

Cite

@article{arxiv.2603.26327,
  title  = {Making Multi-Axis Models Robust to Multiplicative Noise: How, and Why?},
  author = {Bailey Andrew and David R. Westhead and Luisa Cutillo},
  journal= {arXiv preprint arXiv:2603.26327},
  year   = {2026}
}

Comments

9 pages (26 with supplemental), 4 figures (+2 in supplemental), preprint

R2 v1 2026-07-01T11:40:37.927Z