Reproduce “Handling Incomplete Heterogeneous Data using VAEs” 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

Alfredo Nazabal, Pablo M. Olmos, Zoubin Ghahramani, Isabel Valera

HI-VAE imputes missing fields in mixed-type records by learning from observed entries and assigning each field a likelihood suited to its type. In six UCI dataset experiments, entries were removed completely at random. At 20% missingness on Adult, its nominal-field error was 0.304 ± 0.006, versus 0.405 ± 0.002 for mean imputation; it was best on four of six datasets.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.