Machine learning generated predictive model to forecast the dynamic flux distributions of ultra-relativistic electrons

A system comprising:

• a receiver configured to receive a measured low-energy electrons that is less than or equal to 1.5 MeV, wherein the measured low-energy electrons include at least a first and a second set of data, wherein the first set of data is associated with a first electron count rate and wherein the second set of data is associated with a second electron count rate, and wherein the receiver is further configured to receive solar wind data as a third set of data, and wherein the receiver is further configured to receive data associated with observation of higher-electron events with greater than or equal to 2 MeV;
• and
• a processor configured to generate a plurality of machine learning (ML) models based on:
• selection of one or more ML algorithms,
• selection of input data from at least two or more of the first set of data, the second set of data, and the third set of data,
• the data associated with the observation of higher-electron events with greater than or equal to 2 MeV, and
• selection of a window size,
• wherein a first generated ML model of the plurality of ML models is different from a second generated ML model of the plurality of ML models,
• and wherein the processor is further configured to select a subset of ML models from the plurality of ML models to generate an ensemble ML model for predicting high-electron events of greater than or equal to 2 MeV.

A method comprising:

• receiving a plurality of data sets, wherein the plurality of data sets includes a measured low-energy electrons that is less than or equal to 1.5 MeV, and wherein the plurality of data sets further includes data associated with solar wind;
• receiving measured data associated with higher-electron events of greater than or equal to 2 MeV;
• and
• in response to a selection of at least two data sets from the plurality of data sets, and further in response to a selection of one or more machine learning (ML) algorithms from a plurality of ML algorithms, and further in response to a selection of a window size, generating a plurality of ML models based on the selections as an input and the measured data associated with higher electron events of greater than or equal to 2 MeV for predicting high-electron events of greater than or equal to 2 MeV.