<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Ai-Astronomy on Startgaze — Astronomy, Stargazing &amp; Space Science</title><link>https://www.startgaze.com/categories/ai-astronomy/</link><description>Recent content in Ai-Astronomy on Startgaze — Astronomy, Stargazing &amp; Space Science</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Thu, 04 Jun 2026 07:37:18 +0000</lastBuildDate><atom:link href="https://www.startgaze.com/categories/ai-astronomy/index.xml" rel="self" type="application/rss+xml"/><item><title>Rubin Observatory sends millions of alerts per night. Nine ML brokers sort them.</title><link>https://www.startgaze.com/posts/2026-06-04-rubin-observatory-ml-alert-brokers/</link><pubDate>Thu, 04 Jun 2026 07:37:18 +0000</pubDate><guid>https://www.startgaze.com/posts/2026-06-04-rubin-observatory-ml-alert-brokers/</guid><description>Rubin Observatory sent 800,000 sky alerts its first night, scaling toward 10 million. Nine ML broker pipelines filter and classify every one before sunrise.</description></item><item><title>Machine learning is learning to hear inside the Sun</title><link>https://www.startgaze.com/posts/2026-05-21-machine-learning-hear-inside-the-sun/</link><pubDate>Thu, 21 May 2026 07:39:00 +0000</pubDate><guid>https://www.startgaze.com/posts/2026-05-21-machine-learning-hear-inside-the-sun/</guid><description>A Sheffield team ran 30 years of GONG helioseismic data through three ML models. All three predict Solar Cycle 25 bottoms out around 2030–2031.</description></item><item><title>RAVEN found 118 planets in NASA's TESS data — here's how the algorithm works</title><link>https://www.startgaze.com/posts/2026-05-12-raven-ai-finds-118-exoplanets-in-tess-data/</link><pubDate>Tue, 12 May 2026 07:36:53 +0000</pubDate><guid>https://www.startgaze.com/posts/2026-05-12-raven-ai-finds-118-exoplanets-in-tess-data/</guid><description>A Warwick team used gradient-boosted trees and Gaussian processes to validate 118 exoplanets from 2.2 million TESS stars, including rare Neptunian desert worlds.</description></item><item><title>How plate solving works: the algorithm behind every smart telescope</title><link>https://www.startgaze.com/posts/2026-04-25-how-plate-solving-works/</link><pubDate>Sat, 25 Apr 2026 07:35:06 +0000</pubDate><guid>https://www.startgaze.com/posts/2026-04-25-how-plate-solving-works/</guid><description>Plate solving turns a photo of stars into a sky position in about three seconds. Here&amp;#39;s how astrometry.net&amp;#39;s quad-hash algorithm works, and why every smart telescope on the market depends on it.</description></item></channel></rss>