Derived data for: Eruption forecasting of Mt Aso using machine learning on multimodal data with optimized timescales

Derived products (daily prediction indices of single-dataset models for all candidate windows and folds, features selected in each fold, feature matrices of the windows used, feature series of Fig. 9, full list of selected features, tilt corrections, metadata). The raw monitoring data were provided by the Japan Meteorological Agency and NIED Hi-net and are not included. Because the products are derived from JMA data, access is restricted and granted on request.