Communications of the ACM
Artificial Intelligence Helps Stanford Physicists Predict Dangerous Solar Flares
Stanford physicists are bringing artificial intelligence techniques to the challenge of predicting solar flares like this one, captured Jan. 12 by NASA's Solar Dynamics Observatory.
Credit: NASA/SDO and the AIA; EVE; and HMI science teams
Stanford University researchers have automated the analysis of the largest ever set of solar observations to forecast solar flares using data from the Solar Dynamics Observatory (SDO).
The researchers focused on analyzing vector magnetic field data using the Helioseismic Magnetic Imager, a new instrument that collects vector magnetic fields and other observations of the entire sun almost continuously. The researchers process and store the information collected from the SDO, which generates 1.5 terabytes of data a day.
To find new ways to make use of the data, they took an online class on machine learning taught by Stanford professor Andrew Ng. "We had never worked with the machine-learning algorithm before, but after we took the course we thought it would be a good idea to apply it to solar flare forecasting," says Stanford researcher Sebastien Couvidat.
The researchers cataloged flaring and non-flaring regions from a database of more than 2,000 active regions and then characterized those regions by 25 distinct features. They fed the algorithm 70 percent of the data to train it to identify relevant features, and then used it to analyze the remaining 30 percent of the data to test its accuracy in predicting solar flares.
The machine-learning technique confirmed the topology of the magnetic field and the energy stored in the magnetic field are very relevant to predicting solar flares.
From Stanford Report
View Full Article
Abstracts Copyright © 2015 Information Inc., Bethesda, Maryland, USA
No entries found