IBM Granite time-series models enable ensemble forecasting

IBM Research has published a technical post on Hugging Face showing how to use Granite time-series models for ensemble forecasting. The approach combines multiple models to improve prediction accuracy on time-dependent data, addressing a common need in finance, supply chain, and operations. Ensemble methods have long been effective in machine learning; this post shows how to apply them to IBM's new time-series foundation models available on the Hugging Face platform.