“Machine learning” uses clever computer algorithms to go through the huge amounts of data being produced by modern telescopes and spacecraft automatically, flagging those subtle blips that deserve a closer look. I learned about it this month from a postgraduate at the University of Arizona’s Lunar and Planetary Laboratory (LPL) annual student-run departmental conference – this year, conducted online.
In 1975, when I was one of the first students in that then-brand-new department, we didn’t have such colloquia. We also didn’t have that kind of data. We measured our knowledge in bytes, not terabytes. Indeed, our training prepared us for the very different problems of how to measure a signal from a faint body when your electronic detector was not a megapixel CCD image chip, but only a phototube collecting a single dot of light.
02 September 2020, The Tablet
Pilgrims’ progress
Across the Universe
Get Instant Access
Continue Reading
Register for free to read this article in full
Subscribe for unlimited access
From just £30 quarterly
Complete access to all Tablet website content including all premium content.
The full weekly edition in print and digital including our 179 years archive.
PDF version to view on iPad, iPhone or computer.
Already a subscriber? Login