Reproduce “MLJ: A Julia package for composable machine learning” 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
Anthony D. Blaom, Franz Kiraly, Thibaut Lienart, Yiannis Simillides, Diego Arenas, Sebastian J. Vollmer
MLJ represents composite machine-learning workflows as networks of operations rather than fixed sequences. It separates data flow during training from prediction, allowing learned transformations and model parameters to be reused where needed. In a Box–Cox/PCA/ridge example, a learned target transform is inverted after prediction. The paper presents this as an API illustration, not evidence of improved accuracy or speed.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.