Citizen-Science Project to Help NASA Train Machines
Students can become Feature Hunters and help NASA researchers create a machine-learning algorithm to identify surface features in photographs of Earth taken by astronauts in space. NASA has more than 3 million astronaut photos of Earth, many of which have not been examined. Students can help identify glaciers, volcanoes, cities, islands, and more by drawing boxes around them in the photos. Their input will then be used to help train machine-learning models to identify the same geographic features in other NASA astronaut photos of Earth. When the machine-learning models are trained, they will be used to automatically examine millions of photos. An online tutorial explains how students can help identify Earth’s features.
The CanvasMol website has more than 50 three-dimensional interactive rotating models of relatively common molecules such as glucose and fructose. Students can alter each model to show (or not show) bonds, to show (or not show) individual atoms, and to rotate on the X, Y, and Z axes.
The US Department of State has announced STEM Innovations and Global Competence—a free, self-paced online course for US educators. The course focuses on the intersection of STEM subjects and global competence.
The STEM for All Multiplex is an online, free, interactive platform featuring more than 1,600 short videos that showcase federally funded projects aimed at transforming science, technology, engineering, math, and computer science learning.