Comprehensive Time Series Data Analytics for the Prediction of Solar Flares and Eruptions

We report on progress made by our Data Mining Lab at Georgia State University on the interdisciplinary research project funded by the NSF-OAC-1931555 award. We take on two main challenges hindering data-driven solar flare forecasting, which are: (1) extreme class imbalance among solar transient events, and (2) noisy “no-flare” labeling of continuous space weather data.NSF Award # 1931555. Elements: Comprehensive Time Series Data Analytics for the Prediction of Solar Flares and Eruption