Reproduce “Two-stage residual learning for calibrated uncertainty in solar irradiance forecasts” 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

RR Flanagan, E Kasha, PW Talbot, AS Epiny, MR Deinert

The method uses a second LSTM to predict the distribution of errors from a first LSTM’s solar-irradiance forecast, producing a distribution of possible outcomes. Across tested cases, it reduces average deviation between predicted and historical standard deviation by 34.6% for standalone LSTM and 56.5% with Gaussian process regression. Results cover two cities and two forecast horizons, not universal performance.

Summary by alphaXiv.

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