Researchers report that machine learning algorithms trained to classify solar active regions based on whether or not the regions produced an M-class or X-class solar flare presented statistical evidence for previously unknown features of flare-producing active regions, such as the persistence of flare-producing active regions before and after a flare and build-up of electrical currents before a flare.
###
Article #18-20244: "Machine learning reveals systematic accumulation of electric current in lead-up to solar flares," by Dattaraj B. Dhuri, Shravan M. Hanasoge, and Mark C. M. Cheung.
MEDIA CONTACT: Dattaraj Bhalchandra Dhuri, Tata Institute of Fundamental Research, Mumbai, INDIA; tel: +91 9619876816, +91 22 22782679; e-mail: dattaraj.dhuri@tifr.res.in
Journal
Proceedings of the National Academy of Sciences