The data for Precise and Accurate Short-term Forecasting of Solar Energetic Particle Events with Multivariate Time Series Classifiers

This data set contains all the samples (events and non-events) required to run the code for time series classification of predicting SEP events. The "labels.csv" file contains the binary labels of the dataset. A jupyter notebook named "TS_Classifiers_60_min.ipynb" includes the code of our model architecture, including reading the input, training the models, and generating predictions. The model currently runs for short-term (~one hour) prediction of SEP events. Definitions: a. Events are those strong SEPs crossing the 10 pfu threshold in the GOES 10 MeV channels. b. Non-events are weak SEPs with GOES 10 MeV protons below the 10 pfu threshold.