Scientists Use AI To Find Rare Materials
A new machine-learning algorithm can forecast the locations of minerals on Earth and possibly other planets by utilizing patterns in mineral connections. Science and business will greatly benefit from this development as they continue to investigate mineral formations to learn more about the planet’s past and to mine materials for useful products like rechargeable batteries.
An objective that has long been seen as being equally as much an art as it is a science was the focus of a team led by Shaunna Morrison and Anirudh Prabhu. They sought to establish a system for recognizing the occurrence of specific minerals. This procedure has frequently relied on personal expertise and a good dose of serendipity.
A machine learning model from the Mineral Evolution Database predicts mineral occurrences using association rules. The model was tested in the Tecopa basin in the Mojave Desert, predicting locations of geologically important minerals like uraninite alteration, rutherfordine, andersonite.
(With inputs from Shikha Singh)
You need to login in order to Like
