How Bayesian Estimation works in different domains part2(Machine Learning)

<p>In an earlier work, we demonstrated the effectiveness of Bayesian neural networks in estimating the missing line-of-sight velocities of Gaia stars, and published an accompanying catalogue of blind predictions for the line-of-sight velocities of stars in Gaia DR3. These were not merely point predictions, but probability distributions reflecting our state of knowledge about each star. Here, we verify that these predictions were highly accurate: the DR3 measurements were statistically consistent with our prediction distributions, with an approximate error rate of 1.5%. We use this same technique to produce a publicly available catalogue of predictive probability distributions for the 185 million stars up to a G-band magnitude of 17.5 still missing line-of-sight velocities in Gaia DR3. Validation tests demonstrate that the predictions are reliable for stars within approximately 7 kpc from the Sun and with distance precisions better than around 20%. For such stars, the typical prediction uncertainty is 25&ndash;30 km/s. We invite the community to use these radial velocities in analyses of stellar kinematics and dynamics, and give an example of such an application</p> <p><a href="https://medium.com/@monocosmo77/how-bayesian-estimation-works-in-different-domains-part2-machine-learning-8a1ebff6ea16">Website</a></p>