Reproduce “Data Augmentation for Compositional Data: Advancing Predictive Models of the Microbiome” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Elliott Gordon-Rodriguez, Thomas P. Quinn, John P. Cunningham
AugCoDa develops composition-aware augmentation and contrastive representation learning for microbiome classification. It creates training compositions from existing samples through same-class feature mixing and renormalization. Across 12 benchmark tasks, Compositional CutMix achieves the best result on eight, ties on three, and loses one; on type 2 diabetes, DeepCoDa’s test ROC AUC rises from 0.45 to 0.69, a benchmark finding rather than clinical validation.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.