Reproduce “Dual-Context Analog Retrieval for Time Series Forecasting” 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
Jung Min Choi, Ngoc Son Le, Ibram Abdelmalak, Vijaya Krishna Yalavarthi, Lars Schmidt-Thieme
DuoTS forecasts multivariate time series by making a direct prediction, then applying a gated correction to each future segment using recent context and retrieved historical successors. Across eight benchmarks, it has the lowest MSE in 22 of 32 dataset–horizon settings and lowest MAE in 23, versus SRSNet’s nine wins on each metric. It trails on some horizon-average comparisons.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.