Reservoir computing using CVD-MoS2 memtransistor

This dataset contains the complete set of time-series input measurements used to drive the reservoir-computing (RC) pipeline, together with the corresponding one-step prediction outputs computed for every measurement. Each input file provides the raw measurement used as the RC input (typically two numeric columns for V_{GS} and I_{DS} acquired as a time series). For each input, a paired output file provides the model’s one-step-ahead prediction series aligned to the same sample index. Where available, a teacher/target series X(t) is included to reproduce evaluation metrics (RMSE, NRMSE, R^2). File format. Plain ASCII TXT, tab-separated, with a header row; numeric values may use decimal commas or dots and scientific notation (e.g., 3,20157E+0). Typical columns: Inputs (measurement): idx, Vgs (V), Ids (A) Predictions: idx, y_pred_h1 (and, if present, y_true_h1) Intended use. Benchmarking time-multiplexed RC with N=16 virtual nodes (dataset F defaults), optional per-node normalization and lags, and a linear readout. Use together with the companion Software record (Python/Jupyter UI) for parsing, training, metric calculation, and Origin-ready exports.