Enlarge / That’s a lot of mouths to feed. (credit: NASA/Wendy Stenzel ) If you have an emotional attachment to our Solar System’s distinctions, you may want to look away. We’ve found another star system with eight planets, tying our own mark. Oh, and a Google machine-learning algorithm is responsible for the discovery. This is one of two new exoplanets scraped from the massive archive of data from the Kepler space telescope by NASA’s Andrew Vanderburg and Christopher Shallue of the Google AI team. Planets detected by Kepler show up as slight dips in the brightness of a star—the result of the planet passing in front and blocking some of the light. Some planets are more obvious than others, and the goal here was to turn the algorithm loose on digging through past measurements for weak signals that had been missed. Like all machine learning systems, this one was fed measurements from previously identified exoplanets to work out what differentiates real signals from coincidental blips. The researchers say the system emerged with the ability to correctly identify false positives about 96 percent of the time. Read 4 remaining paragraphs | Comments

Continued here:
New discovery pushes star Kepler-90’s menagerie to eight planets