Reproduce “CADENCE: Distributional Next-Transaction Forecasting for Business Cash-Flow Underwriting” 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
Alexander Wu, Jingchao Huang
CADENCE predicts next-payment timing and amount, including no-event probability, for business-counterparty-direction streams. It combines transaction semantics, payment history, and calendar inputs to generate recurrence features. Against production FFT, it improves one-day timing accuracy by 93.3% relatively at matched precision on streams with at least five prior events. Offline underwriting evaluation finds approval rate at fixed loss rate rises 18.2% relatively; this is not causal evidence.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.